A Framework for Sustainable and Fair Demand-Supply Matchmaking Through Auctioning

Abstract

Environmental sustainability and fairness in auction systems are becoming increasingly important as systems evolve with the integration of digital technologies. This paper introduces a novel demand-supply matchmaking (DSM) framework designed to improve fairness and sustainability in auction environments, aligning with the principles of the circular economy. The framework addresses key challenges in supply chain management, such as equitable resource distribution and the reduction of environmental footprints. The framework integrates key aspects of environmental impact assessments, fairness assessments, and behavioral analytics. This enables the simulation of bidder behavior and assessment of auction scenarios. Our simulation results demonstrate that the platform can promote sustainable, fair, and informed auction practices. By comparing our approach with existing tools, we highlight the advantages of using the DSM framework to improve sustainability and fairness in digital marketplaces. This work supports the development of platforms that integrate economic efficiency with environmental responsibility and social equity.

1. Introduction

In recent years, the integration of digital technologies into systems, such as auctioning, has significantly transformed procurement processes, enhancing efficiency while also adding complexity for users like procurement officers and regulatory officers. However, these technological advancements have also highlighted challenges related to environmental sustainability and fairness, which are fundamental to the circular economy framework [1,2]. The circular economy, which focuses on minimizing waste and optimizing resource use, deviates from the traditional linear economy model of “take, make, dispose” [3,4]. The concept is built upon principles such as maintaining the value of products, materials, and resources for as long as possible, minimizing waste, and regenerating natural systems [5]. This approach is increasingly essential for sustainable development, particularly in supply chain management, where resource efficiency is paramount [6,7].

Despite these advancements, traditional auction systems often fall short in addressing these principles, as they are primarily designed for economic efficiency rather than ecological and social equity [8,9]. These systems generally lack the necessary tools to evaluate and optimize economic and environmental performance, leading to inefficiencies and imbalances in resource distribution [10]. Auctions, by their nature, focus on maximizing financial returns, which can sometimes overlook broader social and environmental impacts [11]. For instance, transportation activities linked with auctions contribute significantly to CO₂ emissions, presenting a substantial environmental challenge [12,13]. This makes it difficult for actors like regulatory officers, who need to understand how different regulatory fees affect the market, and procurement officers, who need to define a procurement strategy that aligns well with regulatory settings.

Even with existing advancements in auction systems, there is still a lack of models that address both fairness and environmental sustainability and align with circular economy principles. To address this gap, we introduce a novel demand-supply matchmaking (DSM) model and implement it as part of a novel DSM framework designed to systematically analyze fairness and sustainability in auction environments.

The DSM framework addresses several challenges, including ensuring equitable resource distribution, reducing environmental impacts, and understanding the strategic behavior of bidders. Traditional auctions often result in uneven outcomes where certain participants consistently benefit over others, leading to issues of inequality [14]. By assessing fairness, our DSM framework aims to promote more equitable outcomes among auction participants, ensuring that benefits are shared fairly and transparently [15]. Additionally, the framework incorporates environmental assessments to evaluate and minimize emissions and waste, aligning with global sustainability goals [16,17]. Understanding bidder behavior dynamics is crucial, as strategic bidding often influences auction outcomes significantly. The DSM framework provides insights into these dynamics, allowing for an analysis of how different strategies impact fairness and sustainability, thereby promoting auctions that are both economically efficient and socially responsible [18,19,20].

This paper presents a DSM framework that integrates the DSM model and various other components. The framework includes a platform that provides the technological environment for simulating auction processes and assessing outcomes that derive from the DSM model. This aims to demonstrate the model’s effectiveness through simulations, highlighting its potential to transform auction practices and address existing challenges in the circular economy and supply chain management by promoting equitable resource distribution and reducing environmental footprints [6,15].

Figure 1 illustrates the structure of the DSM framework, showing how the DSM model is combined with mechanisms, tools, and the platform. The DSM model serves as the conceptual foundation, combining environmental impact assessments, fairness assessments, and bidder behavior analysis into a unified logic for analyzing auction scenarios. The operational procedures or algorithms of the model are referred to as mechanisms, guiding how the model’s logic is applied during the platform functions. The platform that hosts this model acts as the technological environment where these concepts are brought to life, allowing users to simulate auction processes and assess outcomes in practice. Within this platform, specific tools are employed to perform key functions; for example, tools for calculating environmental impacts and fairness metrics or for simulating bidder behavior dynamics.

The contributions of this paper can be categorized into three primary areas:

• Problem Analysis and Challenges (Section 3 and Section 4): By addressing shortcomings in existing auction systems, particularly those related to fairness, environmental impacts, and behavioral dynamics, this paper establishes the necessity for a framework that integrates these crucial aspects into the auction process. This analysis identifies where traditional approaches fall short in achieving equitable and sustainable outcomes to fulfill current supply chain and circular economy needs.
• Design and Implementation (Section 5): To address these challenges, the DSM model is designed and implemented within an integrated platform. The model incorporates detailed environmental impact assessments, fairness assessments, and behavioral analytics. These mechanisms are organized into tools that are part of a cohesive and balanced auction DSM framework, capable of addressing the identified shortcomings. Mechanisms within the model guide how these tools and components interact and operate within the platform.
• Evaluation and Results (Section 6): We demonstrate the DSM framework’s practical applicability through its capacity to integrate fairness, environmental impacts, and decision-making processes in auctions. This evaluation highlights the framework’s ability to transform traditional auction systems into more sustainable and equitable frameworks that are fully aligned with the principles of the circular economy.

The results provide detailed evidence of the DSM framework’s ability to increase fairness and reduce CO₂ emissions compared to existing auction methods, showcasing the framework’s potential to transform auction practices [12,16]. These results indicate that the DSM framework can serve as a powerful tool for enhancing both the economic and environmental aspects of auctions, providing tangible benefits for procurement and regulatory officers, as well as other stakeholders. Particularly, understanding how stakeholder engagement shapes these dynamics can reveal new pathways for fostering a sustainability culture in organizations [21].

This paper begins with a literature review that establishes the context for the proposed DSM model by identifying the limitations of current auction systems in terms of fairness and environmental concerns. The review is followed by an exploration of existing tools and technologies, leading to the design and implementation of the DSM framework. The subsequent sections evaluate the framework’s effectiveness through simulations, followed by a discussion on the implications of the findings and potential directions for future research. Finally, this paper concludes with a summary of the key contributions and insights.

By presenting a new approach to auction systems, this research aims to set a new standard for integrating environmental sustainability and fairness into digital marketplaces, ultimately contributing to more responsible and efficient supply chain management and circular economy practices [6,20].

2. Methodology

2.1. Problem Analysis

To identify the key limitations of traditional auction systems, we conducted a comprehensive problem analysis, focusing on three critical areas: fairness, environmental impact, and bidder behavior. This analysis involved reviewing the existing literature and empirical studies on auction practices, particularly those impacting supply chain management and circular economy initiatives. For example, we examined the environmental effects of transportation activities associated with auctions, noting their substantial contributions to CO₂ emissions [12]. Additionally, we reviewed studies highlighting fairness issues, such as unequal resource distribution due to information asymmetry, which disadvantages less-informed bidders. This background informed our selection of fairness and sustainability as the primary objectives for the DSM framework.

2.2. Review of Existing Tools

In this phase, we analyzed current auction platforms to understand their functionalities and limitations. We reviewed tools commonly used for economic efficiency in auction systems and evaluated how they address or fail to address sustainability and fairness. For instance, we assessed platforms primarily focused on maximizing financial returns and identified challenges where environmental impact assessments or fairness metrics were lacking. By comparing these tools, we established the need for an integrated framework that includes environmental and social metrics, offering a more balanced approach to resource allocation within auction systems. This review confirmed the challenges our DSM framework aims to fill, positioning it as a comprehensive alternative to existing tools.

2.3. Design of the DSM Model

Based on insights from the problem analysis and the review of existing tools, we structured the DSM model to integrate fairness, environmental sustainability, and bidder behavior analysis. During the design phase, we focused on creating an adaptable architecture that could evaluate and balance these objectives in auction scenarios. For example, the model was designed with modular components, each addressing a specific metric: fairness indices (e.g., Jain’s Fairness Index), environmental assessments (CO₂ emissions calculations), and bidder behavior simulations.

Each mechanism was designed to function both independently and in conjunction, allowing flexibility in auction setups. We included scenarios that procurement officers and regulatory officers would likely encounter, such as balancing cost with environmental impact in resource allocation. This structured design laid the foundation for the platform’s technical implementation, ensuring each component could be seamlessly integrated.

2.4. Development of the DSM Framework

Building on the findings from the problem analysis and the review of existing tools, we developed the DSM framework, focusing on three core components: fairness metrics, environmental impact metrics, and bidder behavior dynamics. For the fairness component, we integrated Jain’s Fairness Index and percentages to measure equitable outcomes, ensuring fair distribution of resources among participants. For environmental impact, we designed mechanisms to assess CO₂ emissions and waste associated with logistics and transportation, which aligns with circular economy principles. In terms of bidder behavior, we developed a tool to analyze different bidding strategies and their effects on auction outcomes. This design process involved testing various configurations to balance these components without compromising efficiency, as seen in traditional auction systems.

2.5. Platform Implementation

We implemented the DSM model within a digital platform, transforming it into a practical tool for auction-based procurement. This involved creating algorithms and workflows for each component of the model (fairness, environmental impact, and behavior analysis), allowing real-time simulations. For instance, the platform was coded to simulate bidder behavior and calculate fairness metrics dynamically as bids are placed. We employed data-driven methods, such as pre-auction evaluations and environmental assessments, to enrich the decision-making process for procurement and regulatory officers. By consolidating these functions, the platform offers a seamless interface where users can analyze auction outcomes in terms of both sustainability and fairness.

2.6. Evaluation Approach

To validate the DSM framework’s effectiveness, we conducted a series of simulations representing various auction scenarios, each designed to test the framework’s ability to balance fairness, environmental impacts, and cost-effectiveness. For example, we ran simulations comparing aggressive bidding strategies with balanced approaches to evaluate their effects on fairness and sustainability scores. These tests measured the DSM framework’s performance across multiple metrics, such as CO₂ emission reductions and fairness indices, relative to traditional auction systems. By using real-world data on resource distribution and environmental impacts, we demonstrated the DSM framework’s potential to transform auction practices in ways that align with both circular economy principles and sustainable supply chain management.

3. Problem Analysis

Demand-supply matchmaking is complex at a large scale, requiring innovative solutions like the DSM model to address these challenges effectively. We analyze from the perspective of two different roles that represent complementary views on regulatory and market aspects, i.e., the roles of the regulatory officer and the procurement officer.

• Regulatory officers need transparent insights into procurement activities to ensure compliance with fairness and environmental standards, focusing on transparency and accountability to meet ethical and sustainable guidelines.
• Procurement officers face the challenge of balancing competitive bidding with sustainable choices. They must evaluate suppliers responsibly while managing costs. Both roles must consider how different settings and strategies affect fairness, environmental impact, and equitable outcomes, highlighting the intricate challenges of integrating these metrics into large-scale auction processes.

3.1. Fairness in Auction Systems

Fairness in auctions refers to the equitable treatment of all participants, ensuring that everyone has an equal opportunity to succeed based on merit rather than advantage or bias. This includes transparency, accessibility, and impartiality throughout the bidding process, which is especially relevant for regulatory officers and procurement officers working to maintain fair practices.

Traditional auction systems often face criticism for creating unfair outcomes, where participants with more information or resources can gain an advantage, leading to unequal results. This issue is particularly common in sectors like real estate and online marketplaces, where access to information can heavily influence bidding strategies and outcomes [22]. For regulatory officers, such biases challenge efforts to ensure compliance with fairness standards, while procurement officers must navigate these biases to make responsible supplier choices.

Recent studies highlight that information asymmetry and strategic manipulation are persistent issues in auction systems, limiting opportunities for some participants and reducing the perceived integrity of auction systems [22,23]. The reliance on complex algorithms and digital platforms can increase these imbalances, as noted by Kersten, who emphasized the need for auction platforms to incorporate mechanisms that promote transparency and fairness [24]. This focus on fairness is essential for both regulatory and procurement officers to make informed, equitable decisions in auction environments.

3.2. Environmental Impact

Environmental impacts in trade practices are a critical challenge, often overlooked in the pursuit of economic efficiency [1]. Large-scale procurement processes, especially those involving significant logistics, contribute substantially to environmental degradation through resource consumption, waste generation, and greenhouse gas emissions [16]. For regulatory officers, monitoring these impacts is essential for ensuring that procurement aligns with sustainability goals and regulatory standards. Procurement officers are also directly affected, as they must consider these environmental factors when selecting suppliers and making purchasing decisions [25].

The logistics and transportation activities associated with procurement are the primary contributors to greenhouse gas emissions, often conflicting with the sustainability principles that support the circular economy [14]. A recent study by Bocken underscored the urgency of integrating environmental metrics into auction and procurement platforms, emphasizing that sustainable trade practices can significantly reduce ecological footprints [26]. This integration is critical for both regulatory compliance and strategic procurement, as it enables these officers to align trade practices with global sustainability targets. Vijayaraj further advocated for eco-friendly auction designs that minimize environmental impacts, highlighting the importance of sustainable metrics for both regulatory and procurement strategies [27].

3.3. Behavioral Dynamics

Understanding bidder behavior dynamics is essential for predicting auction outcomes and ensuring market efficiency. A bidder’s behavioral pattern is influenced by various factors, including competition, risk tolerance, and information access. These dynamics can lead to strategic manipulation and inefficiencies, complicating the auction process for both regulatory and procurement officers. Regulatory officers must monitor such behavior to prevent unfair advantages, while procurement officers need insights into bidder strategies to make balanced and informed decisions.

Research by Cong demonstrated that strategic bidding often results in outcomes favoring more aggressive or well-informed bidders, which can undermine fairness [28]. Tripathi further explored how behavioral analytics can enhance auction design by providing insights into bidder strategies, promoting outcomes that are both predictable and equitable [29]. This understanding is crucial for developing auction systems that support economic efficiency while upholding social responsibility.

3.4. Circular Economy Challenges

The transition to a circular economy requires auction systems to prioritize sustainability, emphasizing recycling, reuse, and sustainable sourcing. Unfortunately, many auction systems do not align with these principles, contributing to increased waste and resource depletion [7]. This misalignment poses challenges for both regulatory officers, who aim to enforce standards that minimize environmental impact, and procurement officers, who seek suppliers that support circular practices through sustainable product lifecycle management [30].

Stahel highlighted the systemic changes needed in auction models to align with circular economy goals, emphasizing the importance of considering lifecycle impacts in auction design [7]. Recent studies by Moktadir further advocated for integrating circular economy principles into auction platforms, arguing that such integration is essential for fostering sustainable practices and reducing environmental footprints across industries [31].

3.5. Challenges in Supply Chain Management

Current auction systems often create inefficiencies in resource allocation within supply chains, leading to bottlenecks and increased costs. These inefficiencies stem partly from the lack of integration between auction platforms and supply chain management systems, resulting in misaligned procurement, distribution, and inventory strategies [17]. For procurement officers, this misalignment complicates sourcing and resource planning, as they aim to streamline procurement processes to meet demand efficiently. Regulatory officers also face challenges, as inefficient systems can lead to resource waste and excess environmental impacts, making it difficult to ensure compliance with sustainability standards.

Antikainen emphasized that these inefficiencies are a significant barrier to sustainable supply chain management, highlighting the need for systems that support better resource distribution and reduce waste [32]. Wilson further discussed the importance of integrating auction systems with supply chain management to optimize logistics and inventory decisions, enhancing overall efficiency and aligning with sustainable procurement goals [33].

3.6. Interconnections Between Auctions, Supply Chains, and the Circular Economy

The limitations of current auction systems create cascading effects across supply chains and circular economy initiatives. Inefficient auctions can lead to delays, increased waste, and environmental damage, ultimately undermining sustainable resource management goals [5]. For regulatory officers, these inefficiencies complicate efforts to ensure compliance with environmental standards, while procurement officers face challenges in achieving streamlined and sustainable procurement.

The DSM framework addresses these limitations by integrating circular economy principles into auction systems, thereby enhancing supply chain efficiency and sustainability. By incorporating mechanisms that prioritize resource circularity, the DSM framework promotes sustainable outcomes, aligning with global sustainability goals. This aligns with the findings by Shi and Bocken, who highlighted the transformative potential of sustainable auction practices to achieve circular economy objectives [23,26].

4. Existing Tools

In the current landscape of auction systems and supply chain management, various tools and platforms have been developed to increase efficiency, fairness, and sustainability. This section reviews existing solutions, evaluates their capabilities, and identifies the challenges that the proposed DSM framework aims to address.

4.1. Auction Platforms

Auction platforms serve as digital marketplaces where goods and services are bought and sold through bidding processes. These platforms vary widely in their approach to handling fairness, transparency, and environmental considerations.

4.1.1. Traditional Online Auctions

eBay is one of the most prominent online auction platforms, known for its wide accessibility and user-friendly interface. It supports a broad range of products, from consumer goods to specialized items, catering to millions of users worldwide. eBay operates on a dynamic pricing model where the highest bidder wins, offering flexibility and variety in purchasing options. The platform’s accessibility and product variety make it a versatile marketplace, popular among users for its ease of use and comprehensive offerings. While eBay standardizes access to auctions, it does not adequately address fairness strategies by informed bidders or incorporate sustainable practices into its auction framework [34].  

Limitations and challenges:

• Fairness Concerns: Information asymmetry and strategic bidding can lead to unfair advantages, where bidders with more information or faster internet connections can use the system to their advantage. This creates an unequal outcome for less-informed participants.
• Environmental Impact: The platform does not explicitly account for the environmental costs associated with product logistics and transportation, leading to increased emissions and resource use that are not managed or minimized within the platform’s operations.

4.1.2. Auctions and Marketplace

Amazon’s auction-style listings are integrated into its broader marketplace operations, providing users with the option to bid on or buy items directly. The platform is seamlessly connected to Amazon’s extensive logistics and supply chain infrastructure, ensuring an efficient transition from auction to delivery. Amazon’s strong brand reputation and exceptional customer service further enhance user confidence, making it a reliable and trusted platform for online auctions. However, Amazon’s all-inclusive approach is hindered by the limited emphasis on impartial practices and sustainability, resulting in opportunities to improve fairness and reduce environmental impacts [35].  

Limitations and challenges:

• Transparency Issues: Some practices related to seller ratings and product authenticity can affect fairness, as the platform sometimes favors established sellers over new ones, leading to biased auction outcomes.
• Sustainability challenges: There is a limited focus on reducing the environmental impact of auctions, such as emissions from logistics and packaging waste. Amazon’s focus on speed and efficiency often overlooks the environmental costs associated with its extensive logistics network.

4.1.3. Open-Source Auction Platforms

Open-source auction platforms provide flexible and customizable solutions that can be adapted to specific needs and integrated into broader systems.

Auctioneer and FairAuction are two open-source platforms designed to enhance digital auctions. Auctioneer offers extensive customization, supporting various auction types like English and Dutch auctions, making it adaptable to diverse needs [36]. FairAuction focuses on ensuring fairness and transparency in the auction process with tools that promote equitable practices [37]. Both platforms benefit from strong community engagement and allow for tailored solutions for users seeking flexibility. However, they require technical expertise, making them less accessible for non-experts.  

Limitations and challenges:

• Implementation Complexity: These platforms require technical expertise to deploy and customize effectively.
• Limited Support: Open-source tools may lack the support of commercial solutions.

4.2. Environmental Sustainability Tools

With the growing awareness of environmental issues, several tools have emerged to assess and mitigate the environmental impacts of auction and supply chain activities.

4.2.1. CO₂ Calculation Tools for Logistics

These tools, such as DHL’s Carbon Calculator and EcoTransIT World, are designed to assess the carbon footprint of transportation activities within supply chains. These tools offer detailed insights into emissions across various modes of transport, allowing businesses to understand and manage their carbon footprint effectively. Additionally, they provide actionable recommendations for reducing carbon emissions, helping companies develop more sustainable logistics strategies. While useful for logistics, these tools do not address the environmental impact of auction processes, missing opportunities to integrate sustainability into the broader auction framework [38].

Limitations and challenges:

• Limited Integration: These tools often operate as independent tools without direct integration into auction platforms, which limits their ability to provide a broad view of sustainability across the entire auction process.
• Specificity: These tools are primarily designed for logistics and may not account for other environmental impacts, such as those from manufacturing or end-of-life disposal.

4.2.2. Open-Source Environmental Sustainability Tools

OpenLCA is an open-source tool focused on assessing and mitigating the environmental impacts of logistics activities and is specifically designed for lifecycle assessment (LCA). This tool is highly flexible and adaptable to different logistics scenarios. It also benefits from being open source, which allows for extensive customization and community-driven improvements. While OpenLCA provides a flexible and open-source approach to sustainability, it may require significant integration efforts to be effectively used alongside existing auction systems [39].  

Limitations and challenges:

• Integration Challenges: This tool may face difficulties integrating with existing auction systems.
• Limited Support: Community-driven support may not meet all user needs.

4.3. Fairness Tools

4.3.1. Open-Source Fairness Indicators and Algorithms

These tools, such as the Fairlearn toolkit, are designed to measure and enhance fairness in various contexts, including auctions. While these tools provide quantifiable metrics to assess fairness objectively and can be adapted to different auction models, they are not inherently integrated into auction systems, which might require additional customization or integration efforts. Also, their complexity and data dependency limit their applicability and effectiveness in diverse auction environments [40].  

Limitations and challenges:

• Complexity: These tools require extensive data and computational resources to implement effectively, which can be a barrier for smaller organizations or those without significant technical expertise.
• Data Dependency: The effectiveness of these tools depends on the availability and quality of data, which may not always be accessible or reliable.

4.3.2. Open-Source AI Fairness Tools

Aequitas and AI Fairness 360 are open-source tools designed to measure and evaluate the fairness of machine learning models. They provide a range of fairness metrics that can be applied to assess whether a system treats participants equitably, focusing on issues like disparate impact and bias. Their comprehensive metrics provide valuable insights into equity and transparency, making them a great component for platforms aiming to create fair and unbiased environments. However, their direct application in auction systems may require customization or additional integration, increasing their complexity, and data dependency can limit their applicability, especially for organizations lacking the necessary technical expertise [41,42].  

Limitations and challenges:

• Complex Implementation: These tools require substantial data and expertise to deploy effectively.
• Integration Needs: These tools may require additional work to integrate with existing auction platforms, potentially requiring technical expertise and resources.

4.4. Behavioral Analytics

4.4.1. Behavioral Analytics Platforms

These tools, such as Google Analytics and Mixpanel, provide insights into user behavior that can be used to understand bidder strategies in auctions. These platforms offer data-driven insights through real-time analysis of user interactions, helping auctioneers identify bidding patterns and strategies. They are also scalable and capable of handling large datasets across various platforms, making them suitable for extensive auctions. These platforms offer valuable insights but are limited by privacy concerns and a lack of attention to auction contexts, highlighting the need for more targeted behavioral analytics tools in auctions [43].  

Limitations and challenges:

• Privacy Concerns: Handling user data requires careful management to protect privacy, which can be challenging given the increasing focus on data protection and user rights.
• Focus: These platforms are primarily designed for general analytics and may not be tailored specifically for auction environments, leading to challenges in understanding bidder behavior nuances.

4.4.2. Open-Source Behavioral Analytics

Matomo and Open Web Analytics (OWA) are open-source tools that focus on analyzing user interactions, offering a flexible, community-driven approach to understanding user interactions on digital platforms. They provide real-time analysis, helping auctioneers understand bidding patterns and strategies, and benefit from continuous development through community engagement. However, privacy concerns and integration complexities limit their broad applicability [44,45].  

Limitations and challenges:

• Privacy Concerns: Handling user data requires careful management to protect privacy, especially given the increasing focus on data protection. Complexity of Integration: Integration into existing auction platforms may require significant customization and expertise to align analytics with auction-specific metrics.

4.5. Tools Incorporating Multiple Aspects

4.5.1. EnHelix Auction Software

EnHelix 2019 Auction Software is a commercial platform that integrates blockchain technology to increase auction process transparency, security, and efficiency. It is designed primarily for energy and commodities markets, offering tools for environmental impact assessment and compliance with sustainability standards. However, its industry-specific focus and potential implementation complexity may limit its broader applicability [46].  

Limitations and challenges:

• Industry Limitation: This platform is primarily designed for energy and commodities markets, which may limit its applicability to other auction settings or industries.
• Cost: As a commercial solution, it may be costly for smaller enterprises or those with limited budgets.

4.5.2. SAP Ariba

SAP Ariba 2020 is a powerful platform for organizations seeking to integrate sustainability, risk management, and efficient procurement practices. Its comprehensive features support global supply chain operations, making it suitable for large enterprises with complex needs. It provides solutions that can be adapted for auction processes, making it a versatile tool for organizations seeking to increase their procurement strategies. However, its complexity and focus on procurement rather than auctions may necessitate customization and training [47].  

Limitations and challenges:

• Cost Considerations: As a high-end commercial solution, the platform may be expensive for smaller businesses or those with limited procurement needs.
• Limited Auction-Specific Features: While adaptable for auction processes, the platform primarily focuses on procurement and supply chain management, which may require customization for specific auction requirements.

4.6. Shortcomings of Existing Tools

The analysis of current tools, as summarized in

Table 1. Summary of limitations and challenges of auction and related tools.

Tool	Limitations and Challenges
eBay (Traditional Online Auctions)	Fairness issues due to information asymmetry and strategic bidding. Lacks environmental considerations for logistics and transportation.
Amazon Auctions (Auctions and Marketplace)	Fairness concerns related to seller ratings and product authenticity. Limited focus on reducing environmental impacts.
Auctioneer and FairAuction (Open-Source Auction Platforms)	Requires technical expertise for setup and customization. Limited support compared to commercial solutions.
DHL’s Carbon Calculator and EcoTransIT World (CO2 Calculation Tools for Logistics)	Limited integration into auction platforms. Focus on logistics, missing broader environmental impacts.
OpenLCA (Open-Source Environmental Sustainability Tools)	Integration challenges with existing auction systems. Community-driven support may be insufficient for all needs.
Fairlearn toolkit (Open-Source Fairness Indicators and Algorithms)	Requires significant data and computational resources. Effectiveness depends on data availability and quality.
Aequitas and AI Fairness 360 (Open-Source AI Fairness Tools)	Complex implementation, with high data and expertise requirements. May require additional integration work for auction platforms.
Google Analytics and Mixpanel (Behavioral Analytics Platforms)	Privacy concerns related to user data management. General analytics tools not specifically designed for auctions.
Matomo and OWA (Open-Source Behavioral Analytics Tools)	Privacy concerns related to user data handling. Not specifically tailored for auction environments.
EnHelix Auction Software	Primarily designed for specific industries, limiting broader use. Costly for small enterprises with a limited budget.
SAP Ariba	Expensive for smaller businesses. Primarily focused on procurement; may need customization for auctions.

and

Table 2. Comparison of existing tools.

Tool	Fairness	Environmental Impact	Behavioral Analytics	Auctioning	Open Source or Commercial
eBay				✓	Commercial
Amazon Auctions				✓	Commercial
Auctioneer				✓	Open Source
FairAuction	✓			✓	Open Source
DHL’s Carbon Calculator		✓			Commercial
EcoTransIT World		✓			Commercial
OpenLCA		✓			Open Source
Fairlearn Toolkit	✓				Open Source
Aequitas	✓				Open Source
AI Fairness 360	✓				Open Source
Google Analytics			✓		Commercial
Mixpanel			✓		Commercial
Matomo			✓		Open Source
OWA			✓		Open Source
EnHelix	✓	✓		✓	Commercial
SAP Ariba	✓	✓			Commercial

, reveals some important challenges that limit their adoption and usefulness in auction systems. This section highlights these shortcomings and explains their impact.

4.6.1. Lack of Broad Integration

One of the most notable shortcomings of existing tools is their focus on a specific area, such as fairness, environmental impacts, or behavioral analytics, yet they fail to provide a broad approach, as shown in Table 2. For example, tools like Aequitas focus on fairness but do not consider environmental impacts or user behavior. Similarly, tools for calculating carbon emissions like DHL’s Carbon Calculator are great for logistics but do not integrate well with auction platforms. This approach misses the opportunity to create a system that can assess the full environmental impact and fairness of auctions.

4.6.2. High Complexity and Technical Challenges

Many of these tools, especially open-source ones like Auctioneer and FairAuction, require significant technical skills to set up and use, as indicated in Table 1. This can be a significant challenge for smaller organizations or those without dedicated technical teams. This complexity limits the accessibility and broader adoption of these tools, particularly for users seeking straightforward solutions. Additionally, tools that analyze fairness or user behavior often need large amounts of data and powerful computing resources, which are not always available. This makes it harder for many users to use these tools effectively.

4.6.3. Not Specifically Designed for Auction Requirements

Some tools, like Google Analytics and Mixpanel, are designed for general data analysis, not specifically for auctions, as shown in Table 2. This means they might not fully understand or address the unique aspects of bidding behavior in auctions and may need significant customization to be effectively used in auction contexts, which can be resource-demanding, as highlighted in Table 1. Similarly, popular auction platforms like eBay and Amazon Auctions do not focus enough on ensuring fairness or reducing environmental impacts, which are crucial for creating ethical and sustainable auctions.

4.6.4. Cost and Limited Industrial Applicability

Commercial tools like EnHelix Auction Software and SAP Ariba, while offering comprehensive features, are often costly, making them less accessible for smaller enterprises. Additionally, these tools are usually designed for specific industries, like energy or large-scale procurement, making them less flexible and harder to adapt to other types of auctions. As noted in Table 1, these industry-specific tools may not easily adapt to different auction settings without significant customization, further increasing the cost and complexity of their implementation.

4.7. Conclusions

The shortcomings identified in existing tools highlight the need for more integrated, accessible, and auction-specific solutions that can comprehensively address fairness, environmental impacts, and behavioral analytics. Future developments in auction technology should focus on creating tools that are easy to deploy, adaptable across industries, and capable of addressing the full spectrum of ethical and sustainability concerns in auction processes. The DSM framework directly addresses these gaps by combining fairness, environmental impacts, and behavioral analytics into a cohesive platform. By offering tools that are easy to deploy and adaptable across industries, the DSM framework sets novel procurement standards and policies for auction systems, fostering sustainable and equitable practices.

5. Design and Implementation

5.1. Design of the Platform

The platform is designed based on the DSM model to address the key challenges in auction systems. The model consists of two main mechanisms: pre-auction matchmaking and auction simulation. The other mechanisms integrated into the two main mechanisms are fairness assessment, environmental impact assessment, and bidder behavior analysis, as shown in Figure 2. These mechanisms are essential for procurement officers, who require efficient and equitable resource allocation to support responsible purchasing, as well as regulatory officers, who oversee compliance with sustainability and fairness standards.

5.1.1. Pre-Auction Matchmaking of Demanders and Sellers

Before the auction process begins, the model is designed to evaluate potential combinations of demanders (bidders) and sellers (blocks). This process optimizes the strategies for each bidder and suggests scenarios where direct purchases may be more beneficial than auctioning. This process is detailed in [48,49], which provides a comprehensive explanation of the pre-auction part of the model’s evaluation methodology.  

Key Points:

• Optimal Combinations: Evaluate the best matches between demanders and sellers.
• Direct Purchase Opportunities: Identify scenarios where direct purchases may be more advantageous than auctioning.
• Strategic Decision Making: Provides bidders with insights into optimal strategies before the auction process.

Figure 2. The model has several key features, each designed to address a specific aspect of demand-supply matchmaking and auction performance. Open-source model [50].

Figure 2. The model has several key features, each designed to address a specific aspect of demand-supply matchmaking and auction performance. Open-source model [50].

5.1.2. Auction Simulation Process

The model employs the “auction by block” mechanism in this implementation and evaluation. This mechanism auctions each seller’s block sequentially, ensuring efficient and fair resource allocation. This method allows bidders to compete for each block individually, promoting transparent and equitable outcomes.  

Key Points:

• Sequential Bidding: Blocks are auctioned one at a time until all blocks are sold.
• Fairness Mechanisms: Various fairness mechanisms are used to evaluate the equity of auction outcomes.
• Dynamic Behavior: Behavior adjusts based on demand and bidder competition.

5.1.3. Fairness Assessment

Fairness assessment is crucial for ensuring equitable outcomes among bidders. To ensure equitable outcomes, the model incorporates a few fairness mechanisms, including Jain’s Fairness Index, a well-known measure of fairness in resource allocation [51]. These mechanisms are used to evaluate and promote fairness throughout the auction process.  

Key Points:

• Equity Evaluation: Uses fairness mechanisms to assess the equity of auction outcomes.
• Bias Detection: Identifies and mitigates biases in the pre-auction and auction process.
• Transparent Reporting: Provides transparent metrics and reports on fairness.

5.1.4. Environmental Impact Assessment

The environmental impact assessment aims to evaluate and minimize the ecological footprint. This involves mechanisms for carbon emissions and other environmental costs related to logistics, such as transportation and waste. Specifically, waste refers to the difference between the quantity of goods a bidder needs and the quantity they actually receive after the auction. This difference, or excess, is considered waste, which contributes to the environmental footprint of the auction. The model integrates these assessments to promote sustainable practices and align with global sustainability initiatives.

Key Points:

• Carbon Emission Calculations: Evaluate the carbon footprint and suggest charge collection to cover penalties if the footprint threshold is exceeded [52].
• Sustainability Mechanisms: Provide metrics to evaluate and minimize environmental impact.
• Decision Support: Helps organizations make informed decisions that align with sustainability goals.

5.1.5. Behavioral Analysis of Bidders

Behavioral analytics of bidders is essential for creating realistic auction simulations. By modeling different types of bidder behaviors (e.g., aggressive, conservative, and balanced), the platform can better predict and analyze auction outcomes. Defining the behavior templates and their respective strategies allows the simulation to account for diverse bidder actions.  

Key Points:

• Behavior Templates: The behavior templates define the various bidder behaviors, such as aggressive, conservative, and strategic. These templates help simulate different bidder actions and responses during the auction process.
• Dynamic Adjustments: The model allows bidders to dynamically adjust their strategies based on the outcomes of previous auction rounds. Dynamic adjustments ensure bidder behavior remains realistic and adaptable to changing auction conditions.
• Strategic Decision Making: Bidders employ different strategies to maximize their success. In the case of this tool, we calculate the unfulfilled needs of the bidder to change their strategy.

Types of Bidders: The DSM model uses three primary types of bidders: aggressive, conservative, and balanced. Each type of bidder has distinct behaviors and strategies:

• Type A or Aggressive Bidder: These bidders are characterized by high aggressiveness and market price factors. They are willing to place higher bids and take greater risks to win the auction. Their bid likelihood is generally high, and they are less likely to stop bidding, even as prices rise.
• Type B or Balanced Bidders: Strategic bidders balance aggressiveness and conservatism. They adjust their bids based on market trends and unfulfilled needs and aim to make well-informed decisions. Their bid likelihood is balanced, and they stop bidding within a reasonable range to avoid excessive spending.
• Type C or Conservative Bidders: Conservative bidders have lower aggressiveness and market price factors. They are more cautious and place lower bids, aiming to avoid overpaying. Their bid likelihood is moderate, and they are more likely to stop bidding when prices approach their predefined limits.
• Other Types: In this evaluation, we also have types D, E, and F. Each of them is similar to the ones addressed before. For example, type D is nearly identical to type A, with the only difference being their aggressiveness parameter, which is fixed to a value of 0.5, the same as the other types. Similarly, type E is comparable to type B, and type F is comparable to type C. The purpose is to give all bidders of different behavior types in an auction the same starting line when referring to their aggressiveness to bid and see whether that affects the results in the higher bid price.

Behavior Parameters: The behavior of each bidder type is governed by several key parameters, which determine how they interact with the auction process:

• Aggressiveness: This parameter determines how “aggressive” a bidder’s bids are. It effectively scales the bid size. Higher values mean more aggressive bidding, which could result in higher bids and an increased risk of overpaying.
• Market Price Factor: This factor represents the percentage of the market price (price per unit) the bidder is willing to bid. It allows bidders to adjust their bids relative to the market price, making their strategy more flexible and market-aware.
• Stop Bid: This parameter defines the expected price range within which the bidder will stop bidding. It helps prevent overbidding by setting a threshold beyond which the bidder will not go, ensuring that the bids remain within a reasonable range.
• Bid Likelihood: This parameter indicates the likelihood of a bidder placing a bid. It introduces an element of randomness and uncertainty, simulating real-world scenarios where bidders may choose not to bid in certain situations.

5.1.6. Data Analytics and Visualization

Data analytics and visualization are crucial mechanisms of auction systems, providing stakeholders with valuable insights derived from auction data. These mechanisms enable the interpretation of complex datasets, helping make informed decisions and optimize auction strategies. When analyzed and visualized effectively, the files generate data, revealing patterns, trends, and actionable insights.  

Key Points:

• Data Collection: The platform collects data from various pre-auction and auction activities, including bids, bidder behavior, environmental impact metrics, and fairness metrics. These data are stored for further analysis, as shown in Figure 3.
• Data Visualization: The analyzed data are then visualized; in this case, we export the data to an Excel file to make the insights easily interpretable.
• Data Analysis: Once collected, the data are analyzed using statistical methods to uncover patterns and trends. This analysis helps understand bidder behavior, auction outcomes, and the impact of different strategies.

5.2. Implementation Details

The platform unifies the DSM model’s logic by implementing a combination of tools and data-driven methods to address the challenges identified in the design phase. This section details the technical workflows developed to ensure the platform’s functionality, emphasizing features critical for regulatory officers and procurement officers, such as pre-auction matchmaking, auction simulation processes, fairness metrics, environmental impact metrics, and behavioral dynamics, as shown in Figure 4. This state diagram supports compliance, transparency, and strategic decision making in sustainable procurement.

5.2.1. Pre-Auction Matchmaking

The pre-auction matchmaking process is important for optimizing the auction outcomes by evaluating potential combinations of demanders (bidders) and sellers (blocks). This process was implemented using the key functions described below, as shown in Figure 5.

The pre-auction matchmaking process (Algorithm 1) aims to match demanders (bidders) with suppliers (blocks) by first generating all possible combinations of demander–supplier pairs. In Step 1, the find_combinations() function ensures that every bidder is paired with all available suppliers to explore all potential matches. Next, in Step 2, the evaluate_combinations() function assesses each combination by calculating its value, which involves considering the demander’s need, the seller’s supply, and the associated costs. Additionally, the value calculation incorporates metrics like environmental impact (calculate_env_impact) and fairness (calculate_fairness). Finally, these combinations are sorted to prioritize those with the highest value, ensuring that the most beneficial pairs are selected. This approach helps balance economic efficiency with sustainability and fairness in the matching process.

Algorithm 1: Finding and evaluating combinations of demanders and sellers

5.2.2. Auction Simulation Process

The auction process was designed to allocate resources efficiently and fairly. The platform uses an “auction by block” strategy, which was implemented through the functions described below, as shown in Figure 6.

The auction process (Algorithm 2) is conducted for each supplier’s block, ensuring optimal allocation. In Step 1, the conduct_auction_by_blocks() function handles the auction for each block in turn. During Step 2, each bidder places their bids based on their individual strategy—aggressive, conservative, or moderate—using the calculate_bid_amount() function. The bidding process considers different levels of risk and commitment, where aggressive strategies involve higher bids and conservative ones involve lower bids. Once all bids are collected, the highest bid is identified, and the block is allocated to the highest bidder. Finally, in Step 3, the updateVariables() function updates each bidder’s behavior based on the auction round and/or outcome, allowing them to adapt their future strategies dynamically. To accomplish this, the algorithm for the behavior dynamics tool (Section 5.2.5) is called to assign behavior templates and allow dynamic adjustments.

5.2.3. Fairness Metrics Tool

To ensure equitable outcomes, the platform includes a fairness metrics tool, which was implemented as described below, as shown in Figure 7.

Fairness metrics (Algorithm 3) calculate and normalize the fairness measures for the auction outcomes to ensure a balanced allocation of resources. In Step 1, the calculate_jains_fairness_index() function uses Jain’s formula to assess fairness by adjusting each price combination against market or demanded values, providing an initial index of how fairly resources are allocated. In Step 2, the calculate_fairness_percentage() function computes a fairness percentage by comparing the final prices to the demanded prices and accounting for environmental impacts through waste taxes. Finally, in Step 3, the normalize() function scales the fairness values to lie between 0 and 1, ensuring that the calculated fairness metrics are consistent and comparable across all outcomes.

Algorithm 2: Conducting auction by blocks and updating bidder behavior

Algorithm 3: Calculating and normalizing fairness measures

5.2.4. Environmental Impact Metrics Tool

The environmental impact metrics tool was implemented to evaluate and minimize the ecological footprint of auction activities. The key functions are described below, as shown in Figure 8.

Environmental impact metrics (Algorithm 4) calculate and normalize environmental costs and taxes to promote sustainable practices in the auction process. In Step 1, the calculate_waste_taxation() function computes the additional cost for each product based on the percentage of waste generated. A graduated tax rate is applied to encourage reduced waste, thereby fostering more environmentally friendly production and allocation practices. In Step 2, CO₂ taxation is calculated for each bidder by assessing the emissions involved. In Step 3, the calculate_average_distance() function determines the average distance between the bidder and supplier locations using the Haversine formula. Based on these distances, the calculate_CO₂_emissions() function estimates the CO₂ emissions for transportation, considering both truck and plane options, and recommends the mode with lower emissions. After calculating the emissions, Step 3 continues by determining the CO₂ tax using the calculate_CO₂_taxation() function, which applies a graduated rate to reflect the environmental impact of the transport mode used. This layered approach helps ensure that bidders and suppliers are aware of and financially accountable for their environmental impacts, promoting sustainable behavior throughout the auction process. Finally, in Step 4, the normalize() function combines the waste and CO₂ taxation values, scaling them to a value between 0 and 1. This normalization provides a consistent and standardized measure of the environmental impact, making it easier to compare and evaluate environmental sustainability across auction outcomes.

Algorithm 4: Calculating and normalizing environmental impact measures

5.2.5. Behavioral Dynamics Tool

The bidder behavior dynamics tool passes through a set of parameters that influence bidding strategies. This tool was implemented as described below, as shown in Figure 9.

The behavioral dynamics tool (Algorithm 5) demonstrates how bidders adapt their behavior over the course of multiple auction rounds. In Step 1, each bidder is assigned a predefined behavior template—either aggressive, conservative, or balanced—using the assign_behavior_template() function. These templates define how each bidder initially approaches the auction, influencing their bidding strategy and risk tolerance. In Step 2, the bidders are allowed to dynamically adjust their behavior based on the outcomes of the previous rounds. Using the update_behavior_dynamically() function, each bidder evaluates their performance after every round and modifies their strategy accordingly. This approach ensures that bidders learn from their experiences and adjust to changing conditions, simulating a more realistic auction environment where participants can adapt to maximize their chances of success. The dynamic nature of behavior adjustment allows for a more nuanced representation of real-world bidding behavior, where strategies evolve in response to competition and outcomes.

The tool introduced earlier is implemented through a series of calculations that determine each bidder’s actions during the auction. The tool uses predefined behavior templates to assign initial values to each parameter, as shown in

Table 3. Data on behavior templates.

Type	Aggressiveness	Market Price Factor	Stop Bid	Bid Likelihood
Type A	0.8	1.5	2	1.1
Type B	0.6	1.3	1.5	1.1
Type C	0.4	1.1	1.25	1.1
Type D	0.5	1.5	2	1.1
Type E	0.5	1.3	1.5	1.1
Type F	0.5	1.1	1.25	1.1

.

Algorithm 5: Assigning and dynamically adjusting bidder behavior

In each round of the auction, the amount bid by the bidder is calculated using the behavior parameters. If the value of a random number between 0 and 1 (rN) is greater than the bid likelihood (bH), the bidder skips the round to give more variation in the bidder’s behavior. However, to ensure more consistent results, we fix the value of bH to 1.1, as shown in Equation (1). The bid amount (bA) shown in Algorithm 2 is calculated using the aggressiveness (agg), block price (bP), and market price factor (mpF). If the calculated bid amount exceeds the stop bid (sB) multiplied by the block price (bP), the bidder stops bidding for that block, as defined in Equation (2):

$$\mathrm{bA}=\left\{\begin{array}{cc}\mathtt{agg}\times \mathtt{bP}\times \mathtt{mpF}\hfill & \mathrm{if}bH>rN,\hfill \\ \mathtt{skip}\mathtt{round}\hfill & \mathrm{otherwise}.\hfill \end{array}\right.$$

(1)

$$\mathrm{if}sB\times \mathrm{bP}>\mathrm{bA}=\mathtt{Stop\; bidding}$$

(2)

5.2.6. Data Analytics and Visualization

Data analytics and visualization are crucial for interpreting the results of the auction. The implementation involved, as shown in Figure 10, is described below.

The data analytics and visualization tool (Algorithm 6) handles the process of collecting, visualizing, and analyzing data generated before and during the auctions. In Step 1, data are collected from each bidder using the collect_data() function. These data include details on bids, bidder behavior, environmental impact metrics, and fairness indices, ensuring comprehensive coverage of all relevant aspects of the pre-auction and auction processes. In Step 2, the collected data undergo both visualization and analysis. The export_data() function makes the data accessible for further review, exporting it to tools like Excel. Simultaneously, statistical analysis is performed to identify trends and patterns that could support strategic decision making. This dual approach ensures that the collected data are not only presented in an understandable way but also rigorously analyzed to extract actionable insights, ultimately guiding better decision making in future auctions.

Algorithm 6: Collecting, visualizing, and analyzing auction data

6. Evaluation and Results

This study aims to demonstrate the effectiveness of integrating fairness, environmental impacts, and behavioral strategies into auction-based circular economy and supply chain management tools. These tools are essential for procurement officers, who need to optimize resource allocation and make sustainable sourcing decisions, as well as regulatory officers, who ensure that procurement practices align with environmental standards and fairness guidelines. In this section, we present the evaluation and results of our experiments, showing how the proposed framework supports improved decision-making processes, minimizes environmental footprints, and ensures equitable outcomes for all stakeholders.

6.1. Evaluation of the Platform

6.1.1. Evaluation Scenarios

To evaluate the performance and effectiveness of the platform, we tested it across a variety of scenarios that mimic real-world auction conditions. These scenarios were designed to test its performance in terms of fairness, environmental impacts, and bidder behavior.

Scenario 1. 10 Bidders and 14 Blocks: This scenario focuses on a controlled environment where all bidders exhibit the same behavior type (aggressive, conservative, or balanced), highlighting one bidder to facilitate data visualization in the table and charts. The primary aim is to evaluate how uniform behavior types affect auction outcomes, particularly fairness and environmental metrics, when competition is high. This scenario provides insights into the platform’s ability to handle homogeneous bidder strategies and predict outcomes based on pre-auction evaluations.

Roles Involved:

• Procurement officers assess how different bidder behaviors (e.g., conservative vs. aggressive) impact costs and fairness in a highly competitive auction.
• Regulatory officers monitor whether fairness and environmental goals are met uniformly under controlled behavior types.

Scenario 2. Four Bidders and Eight Blocks: This scenario replicates the first scenario but tests the platform in a less competitive environment with fewer bidders and blocks, reducing the intensity of competition. The aim is to analyze whether the platform can maintain fairness and environmental sustainability with lower competition levels. It also evaluates how reduced competition affects the performance of bidder behavior types.

Roles Involved:

• Procurement officers examine whether fairness and sustainability remain achievable with reduced competition, which can simulate niche markets or smaller auctions.
• Regulatory officers verify that sustainability and fairness objectives are not compromised in smaller auction scenarios.

Scenario 3. 10 Bidders, 14 Blocks, and Mixed Behavior Types: This scenario introduces a mix of bidder behaviors (e.g., aggressive, conservative, and balanced), creating a dynamic and complex environment. The goal is to analyze how the interaction of diverse strategies impacts fairness, environmental outcomes, and auction efficiency. This scenario highlights the robustness of the platform in handling real-world complexities. To simplify the data, when visualized in the tables and charts, the bidders were grouped into two categories: normal (types A, B, C, and a combination of them) and moderate (types D, E, F, and a combination of them).

Roles Involved:

• Procurement officers understand how mixed strategies can influence costs, resource allocation, and sustainability outcomes.
• Regulatory officers ensure that mixed strategies do not lead to unfair advantages or compromise environmental goals.

6.1.2. Data Collection Process

The data used in these scenarios were generated using simulated auction environment templates designed to mimic real-world procurement and supply chain conditions. These simulations were based on typical auction datasets, with parameters like demand, supply, waste percentages, and CO₂ emissions. The bidder behaviors (aggressive, conservative, and balanced) were modeled using decision-making rules commonly observed in auction theory and behavioral economics.

Limitations of the Data:

• Real-World Variability: The scenarios do not account for unpredictable real-world factors, such as market volatility, external economic influences, or human biases, that could affect bidding behavior.
• Simplified Assumptions: The bidder behavior models assume rational decision making, which may not fully represent the complexity of human behavior in auctions.
• Dataset Size: The relatively small number of bidders and blocks in the scenarios limits the generalizability of the results to larger, more complex auction systems.
• Static Environmental Metrics: The environmental metrics (e.g., CO₂ emissions and waste percentages) are based on fixed assumptions and may not capture dynamic changes in logistics or supply chain operations.

6.1.3. Evaluation Metrics

To assess the platform’s performance, the metrics described below were applied across all scenarios, as shown in Figure 11.

Fairness Metrics: Fairness was evaluated using Jain’s Fairness Index, which measures how equitably resources are allocated among bidders. Additionally, we looked at fairness percentage metrics, which compare the final pre-auction or auction prices to the initial demands and adjust for environmental impact taxes.

Environmental Impact Metrics: Environmental impacts were measured through metrics such as waste generated (the difference between the amount needed and the amount received by each bidder) and CO₂ emissions related to transportation. These metrics helped assess how well the platform encourages environmentally responsible bidding strategies.

Behavioral Analysis Metrics: Bidder behavior was analyzed by examining the bid success rates, the overall costs incurred by bidders, and how well bidders could adapt their strategies based on auction progress. This was crucial for understanding the platform’s ability to model and predict the outcomes of different bidder behaviors accurately.

6.2. Experimental Platform Setup

Parameters in Tables: The evaluation scenarios are based on simulated auction environments, with the results presented in

Table 4. Results for Bidder 0 in the scenario with 10 bidders × 14 auction blocks.

Type	Need	Supplied	% Waste	Distance Km	% CO2	NEI	Price Total	NF	Weight	Score
Type A	315	632	15	3669.88	20	0.4167	1381.50	0.4714	50/50	0.4440
Type B	315	632	15	3669.88	20	0.4167	860.56	0.4719	50/50	0.4443
Type C	315	632	15	3669.88	20	0.4167	464.34	0.8220	50/50	0.6193
Type D	315	632	15	3669.88	20	0.4167	863.44	0.4719	50/50	0.4443
Type E	315	632	15	3669.88	20	0.4167	717.13	0.7220	50/50	0.5693
Type F	315	632	15	3669.88	20	0.4167	580.42	0.8220	50/50	0.6193
Best Score	315	357	4	1320.04	4	0.8667	386.13	0.7811	50/50	0.8239
Best NEI	315	357	4	1320.04	4	0.8667	386.13	0.7811	50/50	0.8239
Best NF	315	357	4	1320.04	4	0.8667	386.13	0.7811	50/50	0.8239
Pre-Auction	315	632	15	3669.88	20	0.4167	872.16	0.4719	50/50	0.4443

and

Table 5. Results for Bidder 0 in the scenario with 4 bidders × 8 auction blocks.

Type	Need	Supplied	% Waste	Distance Km	% CO2	NEI	Price Total	NF	Weight	Score
Type A	315	1655	30	1884.71	6	0.4	3612.45	0.4621	50/50	0.4311
Type B	315	1655	30	1884.71	6	0.4	2250.26	0.4379	50/50	0.4189
Type C	315	1655	30	1884.71	6	0.4	1214.19	0.7879	50/50	0.5939
Type D	315	1655	30	1884.71	6	0.4	2257.78	0.4379	50/50	0.4189
Type E	315	1655	30	1884.71	6	0.4	1875.21	0.6879	50/50	0.5439
Type F	315	1655	30	1884.71	6	0.4	1517.73	0.7879	50/50	0.5939
Best Score	315	758	15	1884.71	6	0.65	924.00	0.6106	50/50	0.6303
Best NEI	315	758	15	1884.71	6	0.65	924.00	0.6106	50/50	0.6303
Best NF	315	758	15	1884.71	6	0.65	924.00	0.6106	50/50	0.6303
Pre-Auction	315	1655	30	1884.71	6	0.4	2280.59	0.4379	50/50	0.4189

. These tables summarize key parameters for each scenario and their impact on metrics. Need is the quantity of goods that the bidder demands. Supplied is the total quantity supplied by the sellers, which can be one block or a combination of more than one. % Waste and % CO₂ are the extra percentages that would be charged over the total price of the supplies (with a range between 0% and 30%). NEI (normalized environmental impact) is a normalized measure reflecting environmental sustainability. NF (normalized fairness) is a normalized measure indicating fairness in resource distribution. Price Total is the final cost after incorporating all adjustments. Score is a weighted metric combining NEI and NF (50/50 weight in this analysis) to evaluate overall performance.

Chart of 12 Parameters: Figure 12a,b visually compare the results of each behavior type against the best outcome obtained from the pre-auction evaluation by the specific bidder. BestS refers to the best score that Bidder 0 could have obtained by fulfilling the demanded quantity of goods. BestNEI represents the best environmental impact value, indicating the lowest environmental cost achieved. BestNF indicates the best fairness score, reflecting the most equitable allocation. preEva represents the pre-auction outcome for the combination of blocks that Bidder 0 bid on and won, providing a useful benchmark for comparing auction results.

Chart of 13 Parameters: The different prices shown in Figure 13a,b (Price + Discount/Bid, Price + Waste Tax, Price + CO₂ Tax, and Price Total) are the values that affect Jain’s Fairness Index. The parameter Price used in each of the different prices in the chart is the outcome price of the auction in the case of the post-auction calculations and the seller price of the block in the case of the pre-auction calculations. Also, the first bar in the chart changes depending on whether it is a pre-auction or post-auction calculation. In the case of the pre-auction parameter (BestS, BestNEI, BestNF, and preEva), the first price refers to the discount that the bidder can obtain if the combination of blocks has two or more blocks from the same seller, emulating a more realistic market. In the case of the post-auction parameters for all behavior types, the first price considers the winning bid made by Bidder 0. To calculate Price Total, we use the price discount/bid with the percentage taxation of waste and CO₂.

Chart of 14 Parameters: The strategic advantage of conducting a pre-auction evaluation is vividly demonstrated in Figure 14a,b, with a visualization of the Score, NF, and NEI. This evaluation score allows bidders to analyze and identify optimal block combinations before the auction. It also helps the bidder visualize the median of this set of blocks offered for auction and determine whether a good score would be a straightforward approach.

Charts of 15 to 18 Parameters: The last charts in Figure 15, Figure 16, Figure 17 and Figure 18 are mainly divided into two sections: the left side shows behavior types A, B, C, and 1, while the right side shows behavior types D, E, F, and 2, where behavior types 1 and 2 are a combination of the other three respective behavior types on this side.

6.3. Results

This section presents the outcomes from the simulation platform, focusing on the three scenarios outlined in the Evaluation section. The results are analyzed in terms of fairness, environmental impacts, and bidder behavior, providing insights into the framework’s effectiveness and potential areas for improvement.

6.3.1. Scenario 1: 10 Bidders and 14 Blocks

In this scenario, the platform was tested with 10 bidders and 14 blocks, evaluating the outcomes for each behavior type of a single bidder when all bidders followed the same strategy. The key metrics, including fairness, environmental impacts (waste and CO₂ emissions), and price rates, were used to assess performance.

Integrating Fairness and Environmental Metrics

Table 4 shows the results for Scenario 1 with 10 bidders and 14 blocks. Type C (conservative bidder) achieved the highest score, emphasizing the effectiveness of conservative strategies in balancing fairness and environmental sustainability. Specifically, types C and F, which balanced fairness and environmental impacts, achieved a combined score of 0.6193. This suggests that it is possible to achieve both environmental goals and fairness without compromising on either.

The charts in Figure 12a and Figure 14a illustrate how different bidding strategies impacted the outcomes. The analysis challenges the assumption that higher environmental impacts result in lower fairness, as shown by the relatively high fairness score (NF) achieved by type F, despite a high environmental impact (NEI).

Price Analysis

The pricing chart (Figure 13a) reveals how the different bidding behaviors impacted the final costs. For Scenario 1, the aggressive bidding strategy of type A resulted in the highest total price, while conservative bidders managed to maintain a lower total price. This variation directly influenced the fairness index (Jain’s Fairness Index), demonstrating the interconnectedness of economic, environmental, and fairness considerations in auction results.

Key Findings:

• Fairness: The platform demonstrated high fairness across all behavior types, with minor variations. Aggressive bidders (type A) tended to win more blocks but at a higher cost, while conservative bidders (type C) showed lower costs but fewer wins.
• Environmental Impacts: Aggressive bidding led to higher CO₂ emissions and waste due to the larger number of blocks won, highlighting the trade-off between competitiveness and sustainability.
• Bidder Behavior: The pre-auction evaluation closely predicted the auction outcomes, indicating the platform’s strong ability to model and manage bidder behavior effectively.

The results for Scenario 1, as shown in Table 4, indicate that type C (conservative bidder) achieved the highest combined score within this specific scenario. This outcome suggests that, in the context of the DSM framework, conservative bidding strategies were more effective at balancing fairness and environmental sustainability in a highly competitive auction environment. In contrast, the aggressive bidding strategy of type A resulted in higher associated costs, which adversely affected its overall score.

6.3.2. Scenario 2: Four Bidders and Eight Blocks

This scenario examined how the platform performs with fewer bidders and blocks. The results revealed how reduced competition impacts fairness, environmental impacts, and the effectiveness of different bidding strategies.

Integrating Fairness and Environmental Metrics Table 5 presents the results for Scenario 2, with four bidders and eight blocks. In this scenario, the balanced bidder (type B) performed consistently, managing to achieve a good balance between fairness and environmental impacts. The reduced competition allowed for more equitable outcomes, with type F again showing strong performance in balancing both metrics.

The charts in Figure 12b and Figure 14b highlight the impact of different bidding strategies on the outcomes. The strategic advantage of the pre-auction evaluation is clear, as bidders who used this information were able to achieve more competitive bids and better alignment with their objectives.

Price Analysis The pricing chart (Figure 13b) for Scenario 2 shows that even in a less competitive environment, the pricing varied significantly depending on the bidding strategy. Conservative bidders were able to maintain lower costs, but with fewer blocks available, the overall impact on fairness and environmental scores was more pronounced than in Scenario 1. This reinforces the importance of considering both fairness and environmental metrics in auction strategies.

Key Findings:

• Fairness: Fairness metrics remained consistent with Scenario 1 but with slightly reduced variation due to fewer bidders.
• Environmental Impacts: Lower competition led to less aggressive bidding, reducing the overall environmental impact.
• Bidder Behavior: The platform continued to accurately predict outcomes, though the impact of the strategy was less pronounced in a less competitive environment.

In Scenario 2, as shown in Table 5, where there were fewer bidders, type B (Balanced Bidder) performed consistently well. The reduced competition in this scenario led to more equitable outcomes in terms of fairness and environmental metrics, indicating that balanced strategies can be particularly effective under conditions of lower competition within the DSM framework.

6.3.3. Scenario 3: 10 Bidders, 14 Blocks, and Mixed Behavior Types

This scenario assessed the platform’s ability to handle mixed bidder behaviors, comparing outcomes for aggressive, conservative, and balanced strategies within a single auction. The focus was on how mixed strategies influence overall fairness, environmental sustainability, and auction efficiency.

The results revealed that for bidders with behavior types 1 and 2, one bidder (Bidder 7) did not win any blocks during the auction. This is reflected as a low spike in all the charts. Additionally, the metrics Score best, NF best, and NEI best serve as references, indicating the best possible outcomes for each parameter across the different bidders.

In Figure 15a,b, bidders with behavior types C and F achieved outcomes closest to these references. Notably, for behavior types 1 and 2, bidders with more conservative strategies generally performed better, except for one aggressive bidder who secured a favorable outcome by being the last to bid, potentially benefiting from lower prices, as other aggressive bidders had already fulfilled their needs.

Figure 16a,b show that, on average, bidders with behavior types 1 and 2 tended to achieve higher values compared to other behavior types, except for types C and F, which exhibited the best results. Despite this, the other types demonstrated a more balanced average outcome among their respective bidders.

In Figure 17a,b, behavior types 1 and 2 showed better average and more balanced outcomes compared to other behavior types when compared against pre-auction values.

Lastly, the analysis in Figure 18a,b indicates that aggressive bidding strategies, like those of type A, tended to increase total costs. In contrast, balanced or strategic approaches, represented by types C and F, consistently resulted in higher overall scores and better fairness (NF) metrics. This suggests that these strategies are more effective in achieving equitable outcomes by integrating fairness and environmental considerations into the bidding process. Additionally, behavior types 1 and 2 tended to produce more average outcomes across bidders, suggesting they may offer a different but viable strategic approach.

Key Findings:

• Fairness: Bidders with conservative strategies generally led to more favorable outcomes in fairness metrics compared to purely aggressive or balanced strategies. Bidders with mixed behaviors (types 1 and 2) showed more balanced but average outcomes, indicating a viable yet less extreme strategic approach.
• Environmental Impacts: Mixed behaviors introduced more variability in environmental impact, with balanced strategies typically showing better performance.
• Bidder Behavior: The platform successfully managed the complexities of mixed behaviors, demonstrating robustness in diverse auction environments.

In summary, the mixed behavior scenario offered insights into the interactions between different bidding strategies within the DSM framework. The results indicate that aggressive bidders (type A) tended to dominate in securing the most blocks. However, when considering fairness and environmental impact metrics, balanced and conservative bidders (type B and type C) achieved higher overall scores. This outcome underscores the importance of integrating diverse bidding strategies within the DSM framework to achieve more balanced and sustainable auction outcomes.

6.4. Summary of Key Findings

The results demonstrate that the demand-supply matchmaking (DSM) platform successfully integrates fairness and environmental metrics into the auction process, supporting procurement officers and regulatory officers in achieving balanced and sustainable decision making. Specifically, the platform effectively manages multiple objectives, such as maintaining fairness and minimizing environmental impacts, without significantly compromising either. This balanced approach results in auction outcomes that are cost-effective, fair, and environmentally aware, aligning closely with the platform’s goals of strategic and responsible decision making.

6.4.1. Strategic Bidding Advantage

Pre-auction evaluations within the DSM platform provide bidders, including procurement officers, with strategic insights that enable the selection of block combinations that minimize waste and reduce environmental impacts. This proactive approach leads to more competitive bids, lowers overall costs, and improves fairness metrics post-auction. By leveraging these evaluations, procurement officers can align their strategies effectively with both their organizational objectives and the broader sustainability goals of the DSM framework, achieving more favorable and sustainable outcomes.

6.4.2. Impact of Bidder Behavior

The findings from the evaluation of the DSM platform indicate that aggressive bidding strategies, while offering potential short-term gains, are often linked to higher overall costs and lower fairness scores within this auction framework. In contrast, balanced and conservative strategies, especially those informed by pre-auction evaluations, produce more sustainable and equitable outcomes. For regulatory officers, these findings underscore the importance of incorporating fairness and environmental considerations into auction systems, especially in circular economy initiatives and supply chain management.

6.4.3. Mixed Behavior Insights

The analysis of mixed bidding strategies within the DSM platform reveals that aggressive bidders tend to secure a greater number of blocks, while balanced and conservative bidders achieve higher scores in terms of fairness and environmental impacts. For procurement officers, these results highlight the value of incorporating diverse bidding strategies to reach sustainable outcomes using the DSM framework rather than relying solely on aggressive tactics.

6.4.4. Platform Strengths

The DSM platform is effective in maintaining fairness and minimizing environmental impacts across various scenarios involving both homogeneous and mixed bidder behaviors. This adaptability demonstrates its potential for real-world supply chain and market applications, supporting regulatory officers in verifying compliance and procurement officers in achieving sustainable resource allocation. However, trade-offs between fairness, environmental impacts, and competitiveness suggest that further model refinement could enhance its performance, particularly in managing highly aggressive or risk-averse strategies without compromising sustainability goals.

7. Discussion and Future Work

The findings from this study highlight the significant potential of the DSM framework to transform auction systems by integrating fairness and environmental impact assessments. These contributions are critical for addressing contemporary challenges in procurement and regulatory practices, supporting the shift toward sustainable development. However, like any framework, the DSM framework has limitations and potential challenges that must be critically analyzed to ensure its real-world applicability.

Limitations of the DSM framework

• Scalability and Complexity: While the DSM framework performs well in controlled simulations, scaling it to handle real-world auctions with hundreds or thousands of participants may introduce computational and logistical challenges.
• Behavioral Variability: The framework assumes rational and predictable bidder behavior. In real-world settings, bidders may act irrationally or strategically in ways that deviate from modeled behaviors, potentially affecting the accuracy of predictions and outcomes.
• Data Dependency: The reliance on accurate environmental and market data is a key limitation. Incomplete or outdated data could lead to suboptimal outcomes, particularly in assessing environmental impact or fairness metrics.

Broader Implications:

• Policy-Making: The DSM framework provides tools for regulatory officers to design policies that balance environmental impacts with market competitiveness. However, careful calibration is needed to ensure that environmental taxes or fairness adjustments do not excessively increase costs or disrupt market equilibrium. Policies should aim to gradually integrate sustainability metrics to avoid market shock.
• Industry Practices: For businesses, the DSM framework offers a pathway for aligning procurement strategies with sustainability goals. By providing actionable insights into bidder behavior and environmental impacts, the framework encourages more responsible and competitive practices across industries.
• Balancing Prices and Market Viability: The framework’s environmental impact considerations, such as CO₂ and waste taxes, can help incentivize sustainable practices. However, policymakers and industry leaders must collaborate to set tax thresholds that drive change without significantly increasing prices or suppressing market activity.

Future Work

Building on the findings and limitations of this study, several areas warrant further exploration to enhance the DSM framework and its applicability:

• Empirical Validation: While the current study relies on simulations, future research should focus on applying the DSM framework to real-world auction systems. Conducting case studies and analyzing outcomes from live auctions would provide empirical evidence to validate the framework’s effectiveness and refine its parameters.
• Integration of Real-Time Data: Incorporating real-time data, such as live market conditions, environmental metrics, and bidder behavior, would make the DSM framework more dynamic and adaptable. This enhancement would allow procurement and regulatory officers to respond to changing conditions more effectively.
• Advanced Behavioral Analytics: Expanding behavioral models to include factors such as risk aversion, strategic collusion, and incomplete information could improve the framework’s ability to simulate real-world bidder dynamics. Additionally, studying how participants adapt to fairness and environmental taxes over time would provide deeper insights.
• Policy Simulations: Simulating the impact of various policy interventions, such as dynamic tax rates based on environmental impacts, would help policymakers design effective regulations. These simulations should consider different market scenarios to identify optimal strategies that encourage sustainability without harming competitiveness.
• Enhanced Environmental Impact Metrics: Future research should aim to incorporate lifecycle analysis and other comprehensive environmental metrics. This would provide a fuller picture of ecological impacts, from resource extraction to end-of-life disposal, enabling more informed policy and procurement decisions.

8. Conclusions

In this work, we developed a framework that integrates environmental impact assessments, fairness metrics, and behavioral analytics to optimize auction strategies. The main contributions are a problem analysis and identification of the key challenges and shortcomings of existing auction systems, the design and implementation of the DSM framework, and an evaluation demonstrating the benefits of the framework. The key aspects addressed by the DSM framework are as follows:

• Environmental Impact Considerations: By incorporating CO₂ emissions and waste management, the framework aids regulatory officers in enforcing sustainable practices and helps procurement officers minimize the ecological footprint of procurement activities.
• Fairness Assessment: Using Jain’s Fairness Index and other fairness metrics supports equitable outcomes, addressing procurement and regulatory concerns surrounding fair auction practices.
• Behavioral Analytics: The framework’s modeling of various bidder behaviors provides strategic insights, aiding procurement officers in supplier selection and helping regulatory officers ensure ethical bidding.

By addressing these aspects, the DSM framework facilitates more sustainable and equitable market practices, specifically by aligning economic efficiency with environmental management and fairness considerations within auction settings. This focused approach paves the way for targeted innovations, such as integrating real-time data, enhancing environmental metrics, and incorporating advanced behavioral analytics into the platform, further supporting the strategic roles of procurement and regulatory officers.

In conclusion, this study’s results demonstrate that integrating environmental impact assessments, fairness metrics, and behavioral analysis into auction systems, as in the DSM framework, can significantly improve the alignment with circular economy principles and sustainable supply chain goals. This approach leads to more equitable and environmentally responsible auction outcomes while enhancing overall efficiency. Our findings indicate that this framework has the potential to positively influence policymaking and market practices, support circular economy objectives, and contribute to sustainable development, specifically assisting procurement and regulatory officers in their roles. Future research could focus on integrating machine learning for predicting bidder behavior, as well as scaling the platform to accommodate more extensive and diverse market scenarios.