Design Method for Platform-Aggregated Life Cycle Ecosystem

Abstract

The circular economy (CE) is increasingly important in ensuring sustainable development. Although interactions among stakeholders are essential to achieving the CE, existing methods associated with life cycle design do not fully consider the synergies between multiple product life cycles (PLCs). Therefore, this paper proposes a platform-aggregated life cycle ecosystem (PF-LCE). The PF-LCE consists of multiple PLCs along with a platform that enables the exchange of goods and data among different PLCs. We also propose a method for supporting the conceptual design of the PF-LCE. Our method includes a four-step process to help exploit the interactions and synergistic effects occurring across PLCs at the design stage. We then use a simulated case study of PLCs sharing goods and data in the shoe industry, considering scenarios ranging from business-as-usual to full integration of the platform with four service providers. The results demonstrate that the designed PF-LCE delivers an increase in profits to service providers alongside reduced costs for consumers. CO₂ emissions also decrease. Therefore, the design method that we propose has the potential to enhance current implementations of the CE in terms of business revenues, consumer satisfaction, and environmental sustainability.

1. Introduction

The circular economy (CE) [1] is the key concept of an ecosystem that achieves sustainable development in an industry [2,3]. It aims to reduce resource input and waste while maximizing the use of natural resources through cyclical material flow [4,5]. The CE also includes various types of stakeholders and products interacting with each other [6,7]. These interactions among stakeholders contribute to economic and environmental improvements. For instance, industrial symbiosis (IS) [8] involves diverse companies working together to improve resource efficiency and reduce waste through the exchange of materials and by-products [9,10].

To realize businesses aimed at the CE, it is necessary to integrate strategies for delivering value to consumers, maintaining product functionality, and reducing waste generation. To integrate these strategies, we must clarify product life cycles (PLCs) and integrate them with products [11]. Existing design methodologies for PLCs focus on designing a PLC for a single product. These methodologies aim to design stakeholder activities within the life cycle of the CE, including the flows of goods and information among activities. However, as mentioned above, the CE involves multiple products, and consequently, it encompasses multiple PLCs. Therefore, to realize the CE using a life-cycle-based approach, it is necessary to target multiple PLCs simultaneously and facilitate stakeholder interactions across these PLCs.

In this study, we focused on the concept of business ecosystems and aimed to realize a broader CE through a life-cycle-based approach by designing ecosystems composed of multiple PLCs that exchange materials and data via a platform. In the business administration studies context, a business ecosystem is one in which stakeholders interact with each other and generate innovation and competitive advantages [12,13]. These interactions are enabled by a platform that mediates various physical goods and is accessible to a wide range of stakeholders [14]. In this paper, we refer to such ecosystems as platform-aggregated life cycle ecosystems (PF-LCEs). In a PF-LCE, multiple stakeholders interact with each other to innovate value propositions and generate synergies across PLCs. These innovations and synergies improve economic benefits and reduce environmental impact.

Business ecosystems consist of multiple interrelated businesses, and, therefore, the more complex interactions become, the more difficult it is to design each individual business [15]. It has been suggested that one approach for mitigating this complexity is to introduce a conceptual design phase. In this design phase, stakeholders and their interactions are clearly defined within the ecosystem prior to designing individual businesses [16]. Similarly, because a PF-LCE is an ecosystem composed of multiple interacting PLCs, the increased design complexity of each PLC caused by these interactions is a challenge. Therefore, this study aimed to support the conceptual design of a PF-LCE by proposing a method that enables designers to clarify individual PLCs and their interactions in a PF-LCE.

This paper is organized as follows: Section 2 reviews existing research. Section 3 clarifies the structure, definitions, and concepts of PF-LCEs based on the literature review. Section 4 presents a case study of the proposed design method. Section 5 and Section 6 discuss the findings and conclusions.

2. Literature Review

A business ecosystem is defined as “an economic community supported by a foundation of interacting organizations and individuals—the organisms of the business world. This economic community produces goods and services of value to customers, who are themselves members of the ecosystem” [17]. In a business ecosystem, interaction between stakeholders generates innovations that enhance their competitive advantage. The concept of business ecosystems has evolved into other concepts, such as innovation ecosystems and entrepreneurship ecosystems, which are business ecosystems that promote innovation and entrepreneurship [18].

In a business ecosystem, a platform connects multiple organizations, actors, activities, and interfaces and is a source of innovation and dynamism for many products and services [19,20]. The company that leads the design, management, and operation of the ecosystem through the provision of this platform is called the platform leader [21]. For example, IBM has established ecosystems in which it provides CPUs as platforms while third parties supply complementary software [22]. Platform-centered ecosystems require a mutually beneficial relationship between the platform leader and other stakeholders [23,24].

Various definitions and characteristics have been proposed to describe a “platform”. Common components in modular-designed products are called product platforms [25]. The modularization of products makes refurbishing and maintenance more efficient [26] and enables the more flexible expansion of product functions [27]. A platform facilitating interactions among diverse groups, including consumers, companies, and organizations, is referred to as a multi-sided platform (MSP) [28]. Platforms within business ecosystems facilitate interactions among diverse stakeholders. Consequently, they are generally considered MSPs. A key factor in the success of a platform-centered ecosystem is attracting users to the ecosystem by leveraging network effects [29]. A network effect describes the value of a product increasing as the number of users of that product increases [30].

IS aims to reduce environmental impact and resource consumption through the utilization of by-products and industrial waste. In ecosystems pursuing IS, such exchanges of by-products and industrial waste take place among companies from different industries [31]. Even though the introduction of a platform is essential to realize interactions among diverse stakeholders, the existing research on IS does not explicitly consider this aspect.

The structural complexity caused by interactions among various participants within a business ecosystem is an important design challenge [15]. To reduce this complexity, various methods for supporting the conceptual design of the ecosystem have been proposed. These methods include clarifying the participants, the interactions among them, and the platform to enable interaction within the ecosystem [16,32,33] based on the business model canvas [34], as well as a method for structuring value creation mechanisms within the ecosystem elements of the structure [35]. However, existing studies do not explicitly consider environmental impacts within the ecosystem in designing business ecosystems. Because a PF-LCE includes multiple PLCs interacting with each other, it is essential to design the PF-LCE conceptually before designing individual PLCs.

For the establishment of a business ecosystem centered on a platform, one suggested approach is as follows. Initially, a company opens one of its value proposition means (for example, physical products, services, or software), allowing other companies to integrate it into their value delivery processes. Consequently, this mechanism evolves into a platform, with its users becoming participants in the ecosystem [36].

Realizing an ecosystem that balances high economic benefits with low environmental impacts necessitates the combined implementation of various CE strategies. For example, these may include resource-efficient value provision models (such as sharing and rental), the maintenance of product functionality, reductions in resource disposal [37,38], and a Product–Service System (PSS), which is a system for delivering not only products, but also experiences to consumers by combining tangible and intangible services [39].

To implement these strategies as concrete services and products, a methodology for designing a PLC has been proposed [40]. The methodology recognizes that clarifying a PLC is essential to realize the CE. A PLC includes the activities of stakeholders, the interactions among them, and product state changes related to the product type. The methodology also points out that designing a type of product and its PLC in an integrated manner is essential for a PLC with high economic benefits and low environmental impacts. The methodology also models a PLC as a process network. The model consists of life cycle processes—wherein stakeholders induce product state changes (for example, manufacturing, use, and recycling)—and the flow of goods and data among them [41]. However, existing design methodologies primarily target single PLCs, rather than ecosystems that encompass multiple PLCs. Therefore, existing life cycle design methods do not support the design of a broader CE scope. Furthermore, they do not explicitly consider the introduction of the platform, which is essential for interactions among stakeholders across different PLCs. Although clarifying a PLC is essential to the CE, existing research regarding PSSs does not consider clarifying a PLC that is related explicitly to the PSS.

When designing a PLC, it is important to design the activities of stakeholders and products based on the utilization of data with digital technologies. For example, Internet of Things technology collects the condition data of products. The data are analyzed with artificial intelligence technology and used to improve the accuracy of quality control and failure prediction. Blockchain technology enhances the reliability and transparency of data shared among various actors, including a Digital Product Passport [42,43]. Although digital technology and data utilization are essential to realize the CE, the existing research regarding IS does not explicitly consider these technologies and their utilization.

3. Materials and Methods

3.1. Clarifying the Structure and Positioning of a PF-LCE

The literature review shows that clarifying PLCs and treating them as specific design targets is essential for achieving a CE. We position PLCs as fundamental building blocks of the proposed PF-LCE. Each PLC involves stakeholders, including consumers and companies, who exchange goods and data. Stakeholders use these goods and data in their activities, resulting in economic benefits, product state changes, and environmental impacts.

Existing research on business ecosystems and IS has highlighted the importance of diverse stakeholder interactions for increasing economic benefits and reducing environmental impacts; however, studies on PLCs often overlook cooperation across different PLCs. The literature review also identifies platforms as key enablers of such interactions. Thus, we position platforms as another essential building block of the PF-LCE to facilitate stakeholder interactions across PLCs.

To summarize, our proposed PF-LCE design consists of PLCs and a platform, as shown in Figure 1. Stakeholders belonging to different PLCs exchange goods and data through the platform. They use these exchanged goods and data to improve their activities. As a result, the economic benefits of each PLC are improved and the environmental impacts in the ecosystem are reduced, thereby contributing to the realization of a CE.

PF-LCEs have some overlap with business ecosystems, IS, and PSSs, but there are key differences. PF-LCEs and business ecosystems are similar in that they both aim to achieve an economic competitive advantage through stakeholder collaboration. However, a PF-LCE differs in the following two key aspects: it explicitly aims to achieve the CE, and it models the ecosystem as an aggregation of PLCs. In addition, a platform in a business ecosystem is not defined. However, because the platform in a PF-LCE mediates goods and data exchange among diverse stakeholders, the platform in a PF-LCE is classified as an MSP. A PF-LCE and IS are similar in that they both aim to reduce environmental impacts through the exchange of goods between diverse companies. However, a PF-LCE differs from IS by introducing platforms into the ecosystem, by exchanging and using data, and by using exchanged data to provide value to consumers. Additionally, a PF-LCE and PSS are similar in that they both aim to provide value to consumers by integrating diverse goods and services. However, they differ in that a PF-LCE explicitly clarifies the PLCs within the related ecosystem, enabling designers to consider the balance between economic and environmental performance during the design process.

3.2. Policies for Designing a PF-LCE

The literature review shows that although methodologies for designing a single PLC have been proposed, there are no methods for the design concepts of an ecosystem that includes multiple PLCs. Because existing studies on the conceptual design of an ecosystem aim to clarify the participants involved and the interactions among them, it is necessary to clarify the PLCs constituting a PF-LCE and their interactions before designing individual PLCs.

According to the structure described in Section 3.1, a PF-LCE consists of multiple PLCs and a platform. Stakeholders belonging to different PLCs exchange goods and data through this platform. Therefore, as shown in

Table 1. Targets of the conceptual design phase in designing a PF-LCE.

Element	Detail
(1) PLCs	This element represents the PLCs that a PF-LCE encompasses. To clarify how the PLCs interacts with each other, the designers design how the stakeholders and products in each PLC send and utilize the goods and data in their activities.
(2) Platform	This element represents the platform in a PF-LCE. To clarify how the platform acts in mediating the goods and data, the designers design the function and activities of the platform. To mediate the various types of goods and data among various stakeholders, the platform should have functions and spaces for the mediation and be accessible for the stakeholders. Furthermore, the platform should incorporate digital infrastructure (such as software, blockchain, and Internet of Things devices) that enables data mediation.
(3) Goods and data exchanged among PLCs	This element represents the goods and data exchanged among PLCs. The platform (Element (1)) mediates the goods and data.

, our design method for a PF-LCE includes the individual PLCs and a platform as targets, together with the goods and data exchanged among different PLCs. Setting the utilization methods of the exchanged goods and data as design targets helps to clarify the interactions among the PLCs. The participants of a PF-LCE depend on the goods and data that are mediated by the platform. Existing research on digital technologies has suggested that the data available to stakeholders and products within the ecosystem depend on their physical activities. Therefore, in our approach, the designers first identify the PLCs that use the goods mediated through the platform as participants in the ecosystem. Then, they design the exchange and use of data across these PLCs.

In existing research, it has been pointed out that sharing one of the means of value provision owned by a company among multiple companies is a useful approach to determining the participants and interactions within a platform-centered ecosystem. Accordingly, our design method treats one of the means of value provision within a PLC as the platform in the PF-LCE to clarify the participants and interactions within that PF-LCE. To identify the candidates for the platform, our design method begins with specifying a PLC. By redesigning one of the means of value proposition within this PLC as the platform, designers produce a detailed image of the functions and structure of the platform. In this paper, the PLC that is first specified and includes the platform is referred to as the “origin LC”. To mediate various types of physical goods among stakeholders, the platform must provide spaces for storing the goods and functions to support the stakeholders in exchanging these goods via the platform. It must also be accessible to the stakeholders. Furthermore, the platform should incorporate digital infrastructure (such as software, blockchain, and Internet of Things devices) that enables data mediation.

It is necessary to evaluate the results of the conceptual design before proceeding to the detailed design of individual PLCs. A simulation method and a simulator (life cycle simulator with information system simulator: LCS-ISS) are proposed for use with multiple PLCs exchanging goods and data [44]. We apply these methods to evaluate the designed result. To simulate PLCs, the LCS-ISS requires connectivity graphs and an LC-IS model. The connectivity graph describes the mechanical structure of the product as connections between elements. Each element represents the components and parts of products by describing them with a set of attributes. The LC-IS model represents the activities within PLCs and is composed of life cycle flow models. A life cycle flow model consists of life cycle processes and the flows of products, parts, and materials between these life cycle processes. A life cycle process represents activities that are involved in a process that changes the state of a product. Each life cycle process is linked to the elements of the connectivity graphs. The codes use the attributes of the elements to calculate the economic and environmental aspects of the PLCs. The activities are modeled as a series of codes in a programming language. In the LC-IS models, the exchange of materials and data between PLCs is represented as flows between life cycle processes.

3.3. A Method for Designing a PF-LCE

In this section, we propose a design method for the conceptual design of a PF-LCE based on the policies we described in Section 3.2. Figure 2 illustrates the overall processes.

3.3.1. Step 1: Specify the Origin PLC

In this step, the designers set the origin LC to identify candidate platforms for the PF-LCE (Figure 3). First, to identify the candidates for the origin LC, the designers model the PLCs associated with the services and products that a company currently offers or aims to realize in the future, using existing methods such as [41].

Then, the designers select the origin LC from the candidates based on their characteristics. A PF-LCE enhances economic and environmental sustainability through the exchange and use of goods and data across multiple PLCs. To fully leverage these characteristics, it is appropriate to select a PLC with the following attributes as the origin LC: (1) a single PLC in which efforts to improve economic performance or environmental impact (for example, adopting alternative value provision models) yield only limited results and (2) a PLC target product that has traditionally been provided through a product sales model but is now being considered for transition to sharing or rental-based models. By redesigning the product as a platform for sharing among diverse stakeholders, the designers shift its provision model from a product sales model to sharing or rental models.

3.3.2. Step 2: Identify the Platform

In this step, we determine the platform of the PF-LCE (Figure 4). First, to identify candidates for the platform, the designers list the means of value proposition within the origin LC. To realize this, the designers use existing modeling methods such as customer journey maps [45] and service modeling techniques [46]. If there are no means of value proposition that mediate the exchange of goods, the designers return to Step 1.

Next, to compare the candidates, the designers redesign the candidates as platforms. In a PF-LCE, a single platform is used by diverse stakeholders. Therefore, designers need to redesign each platform candidate to be provided through sharing or rental-based models. During this redesign, the designers use existing methods for service design such as service blueprints [47] or the business model canvas [34].

Then, the designers select one of the candidates as the platform of the PF-LCE. To maximize the benefits of goods exchange through physical channels, designers should select a candidate that is used by as many stakeholders as possible and that is capable of mediating diverse types of goods.

3.3.3. Step 3: Determine the PLCs That Participate in the PF-LCE

In this step, the designers determine the PLCs participating in the ecosystem based on the platforms identified in Step 2 (Figure 5). First, to identify potential goods that stakeholders utilize, the designers enumerate the goods that can be mediated by the platform.

Next, to determine the participating PLCs, the designers identify products and stakeholders (such as companies and consumers) that use the listed goods. The PLCs associated with these services and stakeholders become participants in the PF-LCE. If there are no goods that can be mediated by the platform or no PLCs other than the origin LC that can be identified as participants in the PF-LCE, the designers return to Step 2.

Then, to specify how data is used in the activities of stakeholders and products within these PLCs, the designers enumerate the data that can be obtained from these PLCs and design how the data is utilized. Through this step, techniques such as brainstorming [48] and the KJ method [49] support the enumeration of goods.

3.3.4. Step 4: Assess the PF-LCE

In this step, the designers evaluate the design results. First, they decide key performance indicators to evaluate the designed PF-LCE based on the characteristics of the PLCs (for example, models of the value proposition, revenue structure, cost structure, and market growth rate). Then, the designers generate an LC-IS model and connectivity graphs resulting from Step 3 and input the model into the LCS-ISS and execute it. After finishing the simulation, the designers receive numerical results. If the designers are not satisfied with the simulation results, they return to Steps 1–3. The designers use an existing method [44] for modeling and simulating PLCs in this step.

4. Case Study

We conducted a case study to evaluate the effectiveness of the design method proposed in Section 3.3. We aimed to confirm that (1) the designers could design the PF-LCE by using the proposed method and (2) the designed PF-LCE achieved economic and environmental improvements. To conduct the case study, we organized a workshop involving students, university faculty members, and engineers as participants.

Table 2. The conceptual design in the case study.

Element	Detail
(1) PLCs	PLC for commuting shoes: Commuting shoes are provided with a rental model. To collect gait data, the provider of the commuting shoes manufactures sensors for the data. Before consumers purchase flat-foot treatment insoles, running shoes, or climbing shoes, the provider inserts the sensors into the commuting shoes. By using the commuting shoes, the sensors collect the gait data and send it to the platform. After collecting data, the consumers pay a data measurement fee. In addition, flat-foot treatment insoles are able to be inserted into commuting shoes for combined use. PLC for the shoe exchange storage: In the case study, a single shoe box serves as a temporary storage location for the exchange of shoes between consumers and the company. We refer to the shoebox as the “shoe exchange storage” in the case study. The shoe exchange storage consists of multiple shoe storage compartments. A shoe storage compartment stores a pair of shoes. The consumers rent shoe storage compartments based on the number of shoes they own. PLC for flat-foot treatment insoles, climbing shoes, and running shoes: The flat-foot treatment insoles are used by inserting them into the commuting shoes. These insoles and shoes are provided with a product sales model. Before manufacturing them, the providers purchase foot shape and gait data from the providers of the shoe exchange storage and commuting shoes.
(2) Platform	The shoe exchange storage: The shoe exchange storage is installed on the 1st floor of an apartment house and equipped with a scanner to scan the foot shape. Before consumers purchase flat-foot treatment insoles, running shoes, or climbing shoes, the consumers use the scanner to scan their foot shape. After scanning, the shoe exchange storage stores the data and the consumers pay a data measurement fee. To send gait and foot shape data to the providers of flat-foot treatment insoles, climbing shoes, and running shoes, the shoe exchange storage is equipped with a single-board computer. It collects data from sensors and a scanner, stores the data locally, and then transmits it over the internet.
(3) Goods and data exchanged among PLCs	Goods: Commuting shoes, flat-foot treatment insoles, climbing shoes, and running shoes. Data: Requests to use shoe storage compartments, and foot shape and gait patterns of consumers

represents the concept of the PF-LCE resulting from the workshop. Figure 6 and Figure 7 represent the overview of the PF-LCE and the platform inside the PF-LCE.

4.1. Design of a PF-LCE

4.1.1. Step 1: Specify the Origin PLC

At the beginning of the workshop, the participants discussed a wide range of products and services to identify the origin LC. Various candidates such as food, clothing, and daily necessities were proposed in the workshop. The workshop participants focused on product groups in the shoe sector, for which a shift from product sales to other models such as sharing and rental had been attempted but not gained traction. Furthermore, the participants considered that the PLC of commuting shoes had limited potential for further improvement when addressed independently and was suitable for testing the application of cross-PLC collaboration, which is the core concept of our design method. Moreover, shoes are closely tied to everyday consumer use, making it feasible to collect various types of data (for example, the physical condition of consumers and consumer mobility). The workshop participants suggested that this data accessibility allowed for diverse data utilization.

On the basis of this perspective, the workshop participants finally selected “commuting shoes” as the target product of the origin LC. Commuting shoes are used regularly and have recently been offered through rental models. By building a PF-LCE starting from this PLC, they aimed to enhance the competitiveness of the rental service and expand the potential to shift away from the traditional product sales model.

4.1.2. Step 2: Identify the Platform

Next, the workshop participants selected the platform that would serve as the core of interactions within the PF-LCE. To identify platform candidates, they listed goods and services used as means of value proposition in the shoe rental service.

As discussed in Section 3.3, a platform must be capable of mediating various physical goods and accessible to a wide range of stakeholders. On the basis of these criteria, the workshop participants identified shoes, logistics services, the website for renting shoes, and shoe boxes, and redesigned them as platforms. In Japan, people take off their shoes indoors, and there is a product inside the home for temporarily storing shoes. This is referred to as a “shoe box”.

Shoes are used by multiple users. However, owing to limitations in size and shape, the workshop participants determined that shoes were not suitable for meditating diverse items. Although efforts like shared delivery have aimed to enable the joint use of logistics services across companies, the workshop participants identified a key issue: the mobility of delivery vehicles, such as trucks, limits the time and location at which stakeholders—especially consumers—can access them. Consequently, logistics services were considered unsuitable as a platform in the workshop. A website for renting the shoes was also considered unsuitable because the website itself did not have functions and spaces for mediating the physical goods.

For the shoe boxes, the workshop participants proposed a model in which a large, shared shoe box was installed on the first floor of an apartment complex, and consumers rented compartments based on the number of shoes they owned. The workshop participants evaluated that this would enable stable access by an unspecified number of stakeholders and facilitate the mediation of fixed-size items by leveraging the box’s original storage function.

After comparing these options, the workshop participants selected the shoe box as the platform. The single, large, shared shoe box to be installed on the first floor of an apartment complex is hereafter referred to as the “shoe exchange storage”, and the rented space for each consumer as the “shoe storage compartment”.

4.1.3. Step 3: Determine the PLCs That Participate in the PF-LCE

To clarify the PLCs that constituted the PF-LCE, the participants listed the stakeholders and services that could use items mediated through the shoe exchange storage selected in Step 2.

First, the workshop participants discussed the potential goods that could be mediated by the shoe exchange storage, such as groceries and daily necessities. In the workshop, the participants recognized the structural and hygienic limitations of the shoe exchange storage, including dirt, odor, and sanitation issues stemming from the shoes. Therefore, the workshop participants decided to limit the goods mediated by the shoe exchange storage to certain types of shoes.

On the basis of this discussion, the workshop participants identified the following products as promising for mediation. Thus, the PF-LCE included the following PLCs:

• Shoes for flat-foot treatment

These are typically sold as a set with insoles for flat-foot treatment. However, the workshop participants suggested that applying the modular design concept would reduce both the consumer’s payment and the manufacturing cost. In detail, the workshop participants regarded components other than the insole between commuting shoes and flat-foot treatment shoes as the same. Therefore, they designed the commuting shoes so that the insoles for flat-foot treatment were insertable into the commuting shoes to reduce the manufacturing and purchase costs for the shoes for flat-foot treatment. Because the therapeutic aspect requires personalization to each user’s physical traits, the provider of flat-foot treatment insoles supplies the insoles with a product sales model.

• Climbing shoes and running shoes

It is important that these shoes fit the consumer’s physical traits, and, therefore, the service provider of climbing and running shoes personalizes them with a product sales model.

Then, to specify how data were used in the activities of stakeholders and products, the participants listed potential data and discussed how they could be used. As a result of the discussion, the workshop participants listed the following types of data to be used.

• Requests to use shoe storage compartments.
• Foot shape and gait patterns of consumers.
• Condition changes in commuting shoes (for example, dirty and worn).
• Foot odor.

Then, the participants identified how those data were utilized, as follows:

• Requests for shoe storage compartment:

These requests are used to allocate a shoe storage compartment to consumers. If there are insufficient shoe storage compartments, the shoe exchange storage provider manufactures additional shoe storage compartments based on these requests and attaches them to the shoe exchange storage.

• Foot shape and gait data:

These data enable the providers of flat-foot treatment insoles, running shoes, and climbing shoes to manufacture personalized shoes and insoles. Traditionally, consumers must visit physical stores for foot scanning and gait analysis and pay a measurement fee every time they want to buy these shoes and insoles. Visiting stores causes CO₂ emissions from using vehicles. The providers face operational costs for in-store data collection. To resolve these problems, the workshop participants proposed acquiring such data once via physical products and storing it on the platform. To obtain foot shape data, a scanner to scan the consumer’s foot shape data is installed into the shoe exchange storage. Gait data are collected using sensors inserted into commuting shoes. Before the consumers buy either flat-foot treatment insoles, climbing shoes, or running shoes, the consumers use the scanner and sensors. After scanning and collecting, these data are stored in the shared shoe box. Once the data are stored, the providers of flat-foot treatment insoles, climbing shoes, and running shoes buy and receive the stored data over the internet when they manufacture the shoes and insoles. To store and send the data, the shared shoe box incorporates a single-board computer. Consumers pay data measurement fees to the providers of the shoe exchange storage and commuting shoes. Because consumers only need to pay for data acquisition once, rather than each time they purchase shoes and insoles, their total payment is reduced.

• Condition changes in commuting shoes:

While data could automate disposal decisions for rental shoes, the equipment required would likely be expensive, and manual inspections during collection could eliminate the need for such data. Therefore, the workshop participants abandoned the collection of these data.

• Foot odor:

While equipment could detect and notify users about odor, the participants found no additional use cases. Therefore, the workshop participants abandoned the use of these data.

4.1.4. Step 4: Assess PF-LCE

To evaluate the concept, we simulated the designed PF-LCE by using an LCS-ISS. In the case study, the main stakeholders are the providers of each product and consumers. Therefore, to assess economic performance, we selected provider profit and consumer payment amount. To assess environmental impact, we used the CO₂ emissions from each PLC, representing the total resource consumption and environmental burden of the ecosystem. To simulate the PF-LCE, the workshop participants modeled the commuting shoes, running shoes, climbing shoes, the shoe exchange storage, and the PLCs involved using connectivity graphs and an LC-IS model.

Figure 8 and Figure 9 present the connectivity graphs of the shoe exchange storage and the shoes. Each type of shoe has almost the same structure but different attribute values. Only the commuting shoes have the space to insert sensors to measure gait data. The shoe exchange storage consists of shoe storage compartments and a scanner for measuring foot shape data. To identify the renter, each shoe storage compartment is equipped with an authentication panel.

Figure 10 represents the design result. In the model, gait and shoe shape data are sent from the usage processes of commuting shoes and the shoe exchange storage to the manufacturing processes of flat-foot treatment insoles, climbing shoes, and running shoes. Multiple patterns of simulations were executed to assess the design results using the LCS-ISS. The scenarios and simulation scenarios are presented in Section 4.2 and Section 4.3.

4.2. Simulation Scenarios

To evaluate the PF-LCE designed in Section 4.1, we developed six scenarios by varying the participating PLCs and the methods of value proposition. In each scenario, we simulated the profit of the providers, total payment from consumers, and CO₂ emissions to evaluate how goods and data exchanges across different PLCs affected the economic benefits and environmental impacts in each PLC.

We describe the settings of each scenario and the basic calculation methods for economic benefits and environmental impacts. Detailed formulations and settings are provided in the Supplementary Material.

4.2.1. Common Settings

We first describe the common simulation settings applied across all scenarios. The shoe exchange storage is assumed to be installed on the ground floor of an apartment complex. On the basis of a survey by the Ministry of Land, Infrastructure, Transport and Tourism [50], we set the PF-LCE target as a single apartment complex housing 100 families. Using household composition statistics from the national census [51], we set the number of consumers using the shoe exchange storage at 275.

Next, we describe the simulation period. To capture multiple shoe replacement cycles, we set the simulation duration to approximately five years (1050 simulations). Each week is divided into four steps to simulate delivery timing and logistics. According to manufacturer data [52], we set the physical lifespan of each shoe to two years and the value lifespan to one year [53].

In Scenarios 4–6, the providers of commuting shoe rentals and shoe exchange storage sell consumers’ foot shape and gait data to the providers of flat-foot treatment insoles, climbing shoes, and running shoes. By purchasing the data, these providers avoid the cost of measuring the data. For commuting shoe and shoe exchange storage providers, selling these data are a source of revenue, whereas for other providers, it is a cost. If the data fee is too high, the savings from avoiding data measuring costs may be offset. Therefore, we iteratively run the simulation to ensure profitability for all stakeholders. As a result, the data transaction fee is set at JPY 100 per dataset. As with the data price, we decide on the rental price of a pair of commuting shoes and a shoe storage compartment per turn by iteratively running the simulation.

Finally, we describe the settings regarding the shoes. Detailed specifications are provided in the Supplementary Material. We set the specifications of the shoe box and each type of shoe based on manufacturer data and government surveys, as shown in

Table 3. Specifications of the products in each scenario.

Product	Sell or Rental Price (JPY)	Manufacturing Cost (JPY)	CO2 Emissions for Manufacturing (kg-CO2)	CO2 Emissions for Disposal (kg-CO2)
Shoe box (Scenario 1)	12,990 ([56])	11,735 ([56,57])	56.637 ([56,58])	0.22 ([56,58])
Shoe storage space (Scenarios 2~6)	10/step	652 ([56,57])	3.147 ([56,58])	0.012 ([56,58])
A pair of commuting shoes (Scenarios 1–2)	10,953 ([59])	10,330 ([59,60])	9.028 ([61,62])	1.22 ([61,62])
A pair of commuting shoes (Scenarios 3–6)	55/step	10,330 ([59,60])	9.028 ([61,62])	1.22 ([61,62])
A pair of flat-foot treatment shoes (Scenarios 1–3)	13,153 ([59,60])	11,223 ([59,60,63])	9.923 ([61,62,64])	1.341 ([61,62,64])
A pair of flat-foot treatment insoles (Scenarios 4–6)	2200 ([59,60,63])	1893 ([59,60,63])	0.895 ([62,64])	0.121 ([62,64])
A pair of climbing shoes	17,981 ([65])	16,958 ([59,60,65])	15.629 ([60,61,65])	2.11 ([61,62,65])
A pair of running shoes	17,152 ([66])	16,175 ([59,60,66])	4.13 ([60,61,66])	0.559 ([61,62,66])
A sensor for measuring gait data	This product is not for sale or rental	23,625 ([67,68])	3.1 ([67,69])	1.625 ([67,70])
A scanner for measuring foot shape data	This product is not for sale or rental	700,000 ([71])	3.254 ([71,72])	1.71 ([71,72])
A single-board computer for storing and sending the data	This product is not for sale or rental	21,120 ([73])	4.265 ([69,74])	2.235 ([70,74])

. Because commuting shoes are used daily, we assume that all residents in the apartment complex use them. For other types of shoes, we assume that only a portion of consumers purchase them. On the basis of foot shape data [54], 20% of consumers are assumed to purchase flat-foot treatment insoles. According to a survey on leisure [55], 9.35% of consumers purchase running shoes and 5.12% purchase climbing shoes. When consumers purchase flat-foot treatment insoles, climbing shoes, and running shoes, they pay JPY 6600 for data measurement. Each consumer rents shoe storage compartments in the shoe exchange storage corresponding to the number of shoes they own.

Scenario 1

In this scenario, all PLCs operate independently, and no exchange of goods or data across PLCs occurs. Commuting shoes and shoe boxes are provided to consumers through a product sales model. Each household purchases its own shoe box. In real-world shoe personalization services, consumers visit physical stores where staff members collect the necessary data [75]. Therefore, in this scenario, the provider of flat-foot treatment insoles sells the insoles together with dedicated shoes. Before purchasing flat-foot treatment insoles, climbing shoes, or running shoes, consumers must visit physical stores to obtain foot shape and gait data. Accordingly, businesses offering these products incur initial costs for purchasing data acquisition equipment. On the basis of these settings, we calculate the evaluation indicators as follows. Detailed formulas and parameter settings are provided in the Supplementary Material.

• Profits of the providers

Each product is sold independently through a product sales model, and, therefore, we calculate the profit by subtracting the manufacturing, delivery, and data acquisition costs from the purchase price paid by consumers.

• Consumer payment

We calculate the payment as the total amount that consumers pay for purchasing each type of shoe and the shoe box.

• CO₂ emissions

After manufacturing, each shoe and shoe box are delivered to households, used, and eventually disposed of. To calculate the entire CO₂ emissions, we track the number of products manufactured, delivered, and disposed of in each simulation step, and multiply these by the CO₂ emissions per product. In addition, when consumers purchase flat-foot treatment insoles, climbing shoes, and running shoes, CO₂ emissions associated with visiting physical stores are generated.

Scenario 2

In Scenario 2, each consumer rents a number of shoe storage compartments corresponding to the number of shoes they own from a single shoe exchange storage. Consumers also obtain foot shape data using a scanner installed alongside the shoe exchange storage. The installation of the scanner requires an initial investment for the provider of the shoe exchange storage. However, in this scenario, because the PLCs of flat-foot treatment insoles, climbing shoes, and running shoes do not participate in the PF-LCE, consumers must visit physical stores and pay a data measurement fee when purchasing these products. All other settings follow those in Scenario 1.

Accordingly, the calculation methods for the evaluation indicators are based on those of Scenario 1, with the following adjustments. Detailed formulas and parameter settings used in this scenario are provided in the Supplementary Material.

• Profits of the providers

We calculate the profit of the shoe exchange storage provider as the total revenue from renting shoe storage compartments minus the costs of manufacturing and delivering shoe storage compartments and the initial investment in the scanner and the single-board computer.

• Consumer payment

Consumers rent shoe storage compartments based on the number of shoes they own. Thus, total consumer payments are calculated as the sum of purchase costs for shoes and rental fees for shoe storage compartments.

• CO₂ emissions

We calculate the CO₂ emissions related to the shoe exchange storage as the total of emissions from the manufacturing, transportation, and disposal of the shoe storage compartments and the scanner.

Scenario 3

This scenario simulates the changes in economic performance and environmental impact resulting solely from providing commuting shoes through a rental model. Consumers periodically exchange rented shoes and pay fees based on the duration of use. The provider of commuting shoe rents commuting shoes and embedded sensors to collect gait data, which are then sold to manufacturers of flat-foot treatment insoles, climbing shoes, and running shoes. All other settings follow those in Scenario 2.

Accordingly, the calculation methods for the evaluation indicators are based on those of Scenario 2, with the following adjustments. Detailed formulas and parameter settings used in this scenario are provided in the Supplementary Material.

• Profits of the providers

We calculate the profit of the commuting shoe rental provider as the total revenue from rental fees minus the costs of manufacturing and delivering the shoes.

• Consumer payment

Reflecting the change in the provision model, we calculate the consumer payment as the sum of rental fees for commuting shoes and shoe storage compartments and the purchase costs of other shoes.

Scenario 4

In this scenario, flat-foot treatment insoles are delivered to the shoe exchange storage and the provider of the shoe exchange storage inserts the insoles into commuting shoes. As a result, dedicated shoes for flat-foot treatment insoles are not manufactured. Foot shape and gait data are recorded using devices installed in the shoe exchange storage and sensors embedded in the commuting shoes. These data are used to manufacture the flat-foot treatment insoles. Therefore, consumers do not need to visit physical stores to provide their data before purchasing the insoles. All other settings follow those in Scenario 3.

Accordingly, the calculation methods for the evaluation indicators are based on those of Scenario 3, with the following adjustments. Detailed formulas and parameter settings used in this scenario are provided in the Supplementary Material.

• Profits of the providers

We calculate the profit for the flat-foot treatment insole provider as the revenue from insole sales minus the costs of manufacturing, delivery, and purchasing data from the providers of shoe exchange storage and commuting shoes. Because consumers do not visit physical stores to measure data, the provider of flat-foot treatment insoles does not incur initial costs for in-store data measuring equipment.

The sensors inserted in the commuting shoes collect gait data, and, therefore, we calculate the profit of the commuting shoe rental provider as the total revenue from rental fees and data sales minus the costs of manufacturing and delivering the shoes and sensors.

The scanner in the shoe exchange storage scans foot shape data. Therefore, we calculate the profit of the shoe exchange storage provider as the sum of revenue from renting shoe storage compartments and data sales minus the costs of manufacturing and delivering shoe storage compartments and the initial investment in the scanner.

• Consumer payment

Flat-foot treatment insoles are used with commuting shoes, and, therefore, consumers do not pay for shoes dedicated to flat-foot treatment insoles.

• CO₂ emissions

We calculated the CO₂ emissions associated with flat-foot treatment insoles as the sum of emissions from the manufacturing, delivery, and disposal of these insoles. Because consumers do not visit physical stores for data measurement for the insoles, no CO₂ emissions are caused by the visit. However, because the sensors are newly used for collecting gait data before purchasing flat-foot treatment insoles, we calculate the CO₂ emissions needed for the manufacturing, delivery, and disposal of sensors for the data collection.

Scenario 5

In this scenario, consumers obtain foot shape and gait data in the same manner as for flat-foot treatment insoles before purchasing climbing shoes. These data are then used to manufacture personalized climbing shoes. Therefore, consumers do not need to visit a store to provide gait and foot shape data before purchasing climbing shoes. All other settings follow those in Scenario 4.

Accordingly, the calculation methods for the evaluation indicators are based on those of Scenario 4, with the following adjustments. Detailed formulas and parameter settings used in this scenario are provided in the Supplementary Material.

• Profits of the providers

We calculate the profit of the climbing shoe provider as the sum of sales revenue minus the costs of manufacturing, delivery, and purchasing the data from the shoe exchange storage and commuting shoes.

• Consumer payment

Once consumers register their foot shape and gait data using the shoe exchange storage and commuting shoes, the data are stored in the platform. Because the same data are shared with flat-foot treatment insoles and climbing shoes providers, consumers do not need to pay for measuring data multiple times.

• CO₂ emissions

Consumers do not visit physical stores for data measurement for climbing shoes, and, therefore, no CO₂ emissions are caused by the visit. Because the sensors are newly used for collecting gait data before purchasing the climbing shoes, we calculate the CO₂ emissions needed for the manufacturing, delivery, and disposal of sensors for the data collection.

Scenario 6

In this scenario, consumers obtain foot shape and gait data in the same manner as for flat-foot treatment insoles before purchasing running shoes. These data are then used to manufacture personalized running shoes. Therefore, consumers do not need to visit a store to provide gait and foot shape data before purchasing running shoes. All other settings follow those in Scenario 4.

Accordingly, the calculation methods for the evaluation indicators are based on those of Scenario 5, with the following adjustments. Detailed formulas and parameter settings used in this scenario are provided in the Supplementary Material.

• Profits of the providers

We calculate the profit of the running shoe provider as the sum of sales revenue minus the costs of manufacturing, delivery, and purchasing the data from the shoe exchange storage and commuting shoes.

• Consumer payment

Once consumers register their foot shape and gait data using the shoe exchange storage and commuting shoes, the data are stored in the platform. Because the same data are shared with flat-foot treatment insoles, climbing shoes, and running shoes providers, consumers do not need to pay for measuring data multiple times.

• CO₂ emissions

Consumers do not visit physical stores for data measurement for the running shoes, and, therefore, no CO₂ emissions are caused by the visit. Because the sensors are newly used for collecting gait data before purchasing the running shoes, we calculate the CO₂ emissions needed for the manufacturing, delivery, and disposal of sensors for the data collection.

4.3. Results

Figure 11 represents the profit for all providers and the total profit in each scenario. To obtain foot shape and gait data, the shoe exchange storage and the commuting shoe providers paid JPY 1,389,191 and JPY 961,300 each for initial investment. The additional cost for the shoe exchange provider was for purchasing the device and the labor fee for measuring foot shape data. The data acquisition and sales fees for the shoe exchange storage provider were JPY 1,955,964 in total. The additional cost for the commuting shoe provider was for manufacturing and sending sensors to measure gait data. By comparison, the data acquisition and sales fees for the commuting shoe provider were JPY 1,325,345 in total. The profits were approximately JPY 566,773 and JPY 364,045 (approximately 75.3% and 31%) higher in Scenario 6 than in Scenario 1, respectively.

By participating in the PF-LCE, the providers of flat-foot treatment insoles, climbing shoes, and running shoes paid additional costs for purchasing foot shape and gait data. They paid JPY 66,000, JPY 28,600, and JPY 145,600, respectively. By purchasing the data, those providers were allowed to avoid the initial investment in devices required for data acquisition and labor wages. In total, the costs were JPY 2,336,590, JPY 1,827,389, and JPY 4,493,206, respectively. In the business-as-usual scenario, providing dedicated shoes for flat-foot treatment insoles and measuring gait data and foot shape data generated profit. By participating in the PF-LCE, these providers lost that profit. As a result, profit for those providers increased by approximately JPY 210,538, JPY 517,608, and JPY 1,775,000, respectively (approximately 12.3%, 120%, and 53.6% higher in Scenario 6 relative to Scenario 1, respectively).

Figure 12 shows the total amount paid by consumers in each scenario. The introduction of the shoe exchange storage increased the payment by JPY 4,826,367 (comparing Scenarios 1 and 2). However, because other PLCs participated, the payment decreased by JPY 10,949,172 (comparing Scenarios 2 and 6). In Scenario 6, consumers did not need to measure gait and foot shape data, respectively, when purchasing the three types of shoes. This decreased each purchase cost by JPY 6600.

Figure 13 shows the CO₂ emissions for each PLC and the entire PF-LCE in each scenario. Comparing Scenarios 1 and 6, the CO₂ emissions for the PLCs of the shoe box, commuting shoes, flat-foot treatment shoes, climbing shoes, and running shoes were reduced by 23,702, 9564, 4158, 343, and 1795 kg-CO₂, respectively. This corresponded to decreases of 68.7%, 48.2%, 95.8%, 19.2%, and 17.6% relative to the emissions in Scenario 1, respectively.

Renting shoe storage compartments decreased CO₂ emissions (comparing Scenarios 1 and 2). By participating in the PF-LCE, the need for consumers to visit stores to measure foot shape and gait data was eliminated. As a result, the CO₂ emissions caused by these visits were reduced.

In the designed ecosystem, sharing gait and foot shape data online among providers of flat-foot treatment insoles, climbing shoes, and running shoes eliminated the need for repeated store visits for purchases. Moreover, designing the flat-foot treatment insoles to be insertable into the commuting shoes eliminated CO₂ emissions from the shoes for flat-foot treatment insoles. The simulation results show that these reductions exceeded the increase in CO₂ emissions from the manufacturing, delivering, and disposal of the sensors, scanner, and single-board computer for collecting gait and foot shape data.

The above results reveal that the PF-LCE designed in this case study improved both the economic and environmental aspects of each PLC and the entire ecosystem.

5. Discussion

Although the proposed PF-LCE shares many similarities with business ecosystems, it differs in its explicit consideration of environmental aspects and the clear representation of PLCs. These differences enable environmental factors to be incorporated into the design process—an aspect typically overlooked in the conventional business ecosystem context.

Similarly, while a PF-LCE and IS have several commonalities, the PF-LCE introduces key distinctions through the explicit use of platforms and the exchange of data among stakeholders. The integration of the platform helps to clarify how interactions among stakeholders are enabled. Furthermore, by incorporating explicit design for data exchange, a PF-LCE allows designers to improve and transform stakeholder activity through digital technologies and data utilization.

The design method proposed in this study supports the conceptual design of such PF-LCEs. Traditional life cycle design approaches are focused on a single PLC, making PF-LCE design unfeasible. By contrast, our method targets the entire PF-LCE, thereby contributing to its practical realization. To clarify the concept of an ecosystem that consists of multiple PLCs, our design method begins with designing a PLC and treats a means of value proposition in the PLC as the platform.

The simulation results from the case study suggest that the PF-LCE, which was designed and constructed with our method to enable the exchange of goods and data across PLCs, improved the economic and environmental performance of each PLC. In the PF-LCE, providers of commuting shoes and the shoe exchange storage faced additional costs for participating in the PF-LCE. Other providers avoided the upfront costs required in the business-as-usual scenario by participating in the PF-LCE. This would allow them to use that capital as initial capital for other businesses. In relation to the environmental aspect, the additional environmental burden required for each provider to participate in the PF-LCE and build new value delivery processes was outweighed by the reduction in environmental impact achieved through the sharing of parts and data.

In the case study, the workshop participants designed a PF-LCE based on the unique usage patterns of shoes and shoe boxes in Japan. However, the design principles proposed in this study, the LC-IS model, and the LCS-ISS are not limited to the Japanese context or specific PLCs and are, therefore, applicable to diverse PLCs across industries. Moreover, the economic and environmental performance of each PLC improved as more PLCs participated in the PF-LCE. This result suggests that the construction of a PF-LCE generates positive network effects. Meanwhile, as the number of PLCs included in a PF-LCE increases, the corresponding LC-IS models become more complex, which, in turn, increases the difficulty of evaluating designed PF-LCEs.

In the designed PF-LCE, sensors and scanners capture data that is transmitted via the internet, and flat-foot treatment insoles are physically embedded in commuting shoes. These features serve as mechanisms for utilizing shared data and physical components. This result suggests that our design method enables designers to design not only platforms, but also mechanisms for sharing and utilizing data and physical parts within ecosystems. Sharing data and parts is increasingly being discussed in more complex and technologically advanced industries—for example, through modular design [76] and Catena-X [77] type initiatives in the automotive sector. These observations suggest that the proposed method serves as an approach for sharing and utilizing data and parts in a wide range of industries, including those with greater product complexity and environmental impact.

Despite these positive findings, the exchange of goods and data among providers in a PF-LCE involves trade-offs. For example, in the case study, data sales generated revenue for the shoe exchange storage and commuting shoe providers while representing a cost for other providers. Therefore, variables related to these trade-offs (for example, the price of data) must be adjusted to ensure that the economic relationship among providers results in a win–win outcome. Achieving this requires iterative simulation.

In addition to the case study, we conducted supplementary interviews with a real-world service designer regarding the proposed design method and simulation results. The following responses were obtained. The designer noted that basing the ecosystem on mediated physical goods made the physical activities within the ecosystem more concrete, which, from his experience, helped him more easily generate ideas for data utilization. However, he expressed concerns about the security risks associated with data sharing through platforms. To promote data sharing, it is essential to protect data rights [78]. In addition to security, the traceability and reliability of data are crucial for data sharing [79].

To ensure secure and traceable data among diverse stakeholders, it is essential to adopt standardized frameworks and incorporate advanced digital technologies. Examples include the Digital Product Passport initiative and blockchain-based software, which enhance data traceability, integrity, and access control. While this study focused primarily on the conceptual and structural design of PF-LCEs, we recognize that the implementation of the platform must integrate such standards and technologies during the implementation or development phase of PF-LCEs.

6. Conclusions

In this paper, with the aim of achieving a CE that targets a wider range of ecosystems, we proposed a method for the conceptual design of a PF-LCE. The PF-LCE integrated the concept of the business ecosystem, platform, and PLCs. The core of the design method was treating a channel between consumers and service providers in the origin PLC as a platform and designing the exchanges among the PLCs on the basis of that platform.

A case study for designing a PF-LCE with four types of shoes and shoe boxes was evaluated. In the PF-LCE, the sharing of parts and data generated synergistic effects across PCLs. The designed PF-LCE improved profits and reduced environmental impacts compared with the business-as-usual case. Through the case study, we confirmed that the proposed design method supports designers in the conceptual design of a PF-LCE and the evaluation of PLCs. However, to facilitate data sharing among different stakeholders, it is essential to ensure security and establish common standards.