Next Article in Journal
Accounting for the Different Propagation Velocities for the Focusing and Time–Depth Conversion in a Layered Medium
Previous Article in Journal
Cyclic Performance of Prefabricated Bridge Piers with Concrete-Filled Steel Tubes and Improved Bracing Connection Detail
Previous Article in Special Issue
A Tent-Lévy-Based Seagull Optimization Algorithm for the Multi-UAV Collaborative Task Allocation Problem
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Editorial

AI Technologies for Collaborative and Service Robots

by
Giovanni Boschetti
1,2,*,†,
Matteo Bottin
1,† and
Riccardo Minto
1,†
1
Department of Industrial Engineering, University of Padova, 35131 Padova, Italy
2
Department of Information Engineering, University of Padova, 35131 Padova, Italy
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Appl. Sci. 2024, 14(24), 11811; https://doi.org/10.3390/app142411811
Submission received: 4 December 2024 / Accepted: 12 December 2024 / Published: 17 December 2024
(This article belongs to the Special Issue AI Technologies for Collaborative and Service Robots)

1. Introduction

In the last few years, Artificial Intelligence (AI) has become increasingly popular, even in the consumer field. This improvement in technology and the increased popularity of multiple platforms have become appealing also for collaborative and service scenarios, in which human behavior may be unpredictable.
According to the International Organization for Standardization, the service robot is a “robot in personal use or professional use that performs useful tasks for humans or equipment” [1], also specifying that “tasks in personal use include handling or serving of items, transportation, physical support, providing guidance or information, grooming, cooking and food handling, and cleaning”, and “tasks in professional use include inspection, surveillance, handling of items, person transportation, providing guidance or information, cooking and food handling, and cleaning”. As a result, service robots can be employed in consumer and professional applications, widening both the appeal and the challenges the robot must face.
The International Organization for Standardization does not provide a specific definition of a collaborative robot. Instead, the standard [1] firstly defines “robot” and then defines “collaboration” as “operation by purposely designed robots and person working within the same space”. It has also become common to define the degree of collaboration based on the amount of interaction between humans and robots within the workspace [2].
Under the framework of Industry 5.0, human wellness is the key aspect to consider when designing robot interactions. Since the robot’s motion may not be predictable by the human operator, the influence of his/her stress on the system performance may become relevant [3,4]. As a result, the robot’s behavior must be designed and adapted according to the specific human needs.
Artificial Intelligence has been employed extensively in collaborative robotics in the last years [5]. However, the interest in AI has yielded novel results which have been published in the Special Issue entitled “AI Technologies for Collaborative and Service Robots”. In the following section, an overview of the published papers will be provided, highlighting the novelty of the new publications.

2. An Overview of Published Articles

Service robots can be employed as a single robot within a productive plant, or, more commonly, as a fleet of robots managed by a central unit. A recent innovative study [6] aims at optimizing unmanned aerial vehicle (UAV) systems, in which the employment of multiple UAVs enhances production efficiency and eases the execution of specific operations. In this work, the UAV application is in the agriculture field; however, the novel approach can be applied to other fields, such as aerospace and military domains. The main contributions of this study are related to establishing a single-objective optimization model for the multi-constrained multi-UAV collaborative task allocation problem (MCMUAVCTAP) and to proposing the Tent–Lévy Improved Seagull Optimization Algorithm (TLISOA) to address the MCMUAVCTAP. The MCMUAVCTAP is an allocation problem in which the fleet manager (i.e., the one assigning the tasks to the single UAVs) has to assign multiple UAVs to collaboratively perform various tasks while adhering to multiple constraints to minimize overall system costs. Simulation experiments in agricultural contexts demonstrate that the algorithm enhances the response speed and efficiency of multi-UAV systems. Moreover, the TLISOA has been compared to five other algorithms (SOA, CAM-GA, A-QCDPSO, PSO-AWOA, and CESMA), demonstrating its efficiency under many performance indexes.
In the field of collaborative robotics, a novel study [7] aims at improving collision avoidance methods. In particular, the study integrates the geometry-aware singularity avoidance cost with the Fast Marching Tree (FMT). The former is a method that employs Riemann distance to measure the proximity of the actual robot configuration with its singular configurations; the latter is a combination of Rapidly Exploring Random Tree Star (RRT*) and the probabilistic roadmap method (PRM), which are used in obstacle avoidance methods. As a result, the proposed method aims to avoid obstacles and singularities at the same time, enhancing the capability of the single algorithm. The algorithm operates with a two-stage strategy: firstly, a geometric envelope method for both the robot arm and the obstacles is used for collision detection; then, a quasi-uniform cubic B-spline curve is employed for path optimization. The optimization is performed in Cartesian space. Algorithm results show the ability to simultaneously achieve good performance and acceptable computational complexity, both in simulated and real-world validation experiments. Again, the method is compared with existing methods to show its strengths.
One of the problems of traditional robotics usually tackled by collaborative systems is the difficulty in programming [8]. Indeed, the modern industrial scenario requires frequent changes to the robot programming to adapt the robot to the rapid changes in the production process. The programming should feature both the traditional robot tasks, such as inspection and pick and place, but also the presence of collaborative tasks, which must consider both the different collaboration modes as intended by the standards [9], but also the different areas of the workspace, where collaboration may or may not occur. These problems are addressed in [10], where the authors present a novel version of their model language Robot Task Modeling and Notation (RTMN), which is based on the Business Process Modeling Notation (BPMN), which is a graphical representation of business processes. The model allows for the intuitive modeling and programming of robotic processes, and the new version includes four human–robot collaboration tasks, as considered by the international standard, requirements, and key performance indexes (KPIs) to evaluate the performance of the robotic workcell. The approach was tested with qualitative and quantitative evaluations on 82 subjects by means of questionnaires and interviews, showing that the interface is intuitive and easy to understand, and the subjects showed interest in using the system.
An alternative approach to simplify robot programming is based on Learning from Demonstration (LfD), i.e., the ability of robots to replicate tasks based on human demonstrations. In this regard, a novel study [11] implements Generative Adversarial Networks (GANs) to efficiently capture the distribution of expert demonstration trajectories. The implementation is further improved by adopting additional loss functions which enhance convergence performance. The method allows researchers to achieve trajectory generalization by adjusting the trajectory through affine transformations. The performance of the algorithm is employed in an experiment in which a UR5 collaborative robotic arm has to perform some trajectories derived from the Lasa dataset. This dataset comprises a collection of 2D handwriting motions recorded from a tablet’s personal computer. To further show the validity of the method, the Lasa dataset is modified with some noise, and the result is assessed against the original data. The method is able to adeptly learn from demonstration trajectories and generate new trajectories that closely mimic the learned skills.
Autonomous vehicles are often associated with AGVs (Autonomous Guided Vehicles) or AMRs (Autonomous Mobile Robots). However, the rise of electric vehicles, autonomous driving, and valet parking technologies has increased the need for autonomous tasks. One of these tasks is the automatic charging of electric vehicles, which has been addressed with an industrial robot with the focus of improving safety [12]. The proposed method is based on a refined collision classification method based on the light gradient boosting machine (LightGBM) handling the environmental noise via an empirical mode decomposition (EMD). The target of the research is to maintain model performance while lowering computational costs. The latter is possible thanks to the LightGBM. Experimental tests are performed with an Aubo-i5 holding the charger. The end effector is flexible and is also equipped with an Inertial Measurement Unit (IMU). To train the model, a set of 289 collision points has been defined. Such points are the positions in which there is enough misalignment between the charger and the charging port that unwanted collision occurs. Moreover, angular deviations in the position misalignment are considered as well. An extensive study of the properties of the sensor’s acquired data is performed, highlighting the features that better describe the presence of collision. Classification accuracy is very high even for low-time steps, outperforming other algorithms.

3. Conclusions

The research manuscript presented in this Special Issue contributed to the development of improved and more resilient approaches for both service robots and collaborative ones. The focus on robot programming, by introducing new frameworks to simplify robot programming considering both traditional and robotic tasks or new neural networks to improve the ability of robots to learn the tasks from human demonstrators, will allow companies to easily adapt robot behavior to changes in the production process, which is particularly necessary for small and medium-sized enterprises.

Author Contributions

Writing—original draft preparation, M.B.; writing—review and editing, G.B. and R.M. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. ISO 8373:2021(en); Robotics—Vocabulary. International Organization for Standardization: Geneva, Switzerland, 2021.
  2. Bauer, W.; Bender, M.; Braun, M.; Rally, P.; Scholtz, O. Lightweight Robots in Manual Assembly—Best to Start Simply; Frauenhofer-Institut für Arbeitswirtschaft und Organisation IAO: Stuttgart, Germany, 2016; Volume 1. [Google Scholar]
  3. Prassida, G.F.; Asfari, U. A conceptual model for the acceptance of collaborative robots in industry 5.0. Procedia Comput. Sci. 2022, 197, 61–67. [Google Scholar] [CrossRef]
  4. Faccio, M.; Granata, I.; Menini, A.; Milanese, M.; Rossato, C.; Bottin, M.; Minto, R.; Pluchino, P.; Gamberini, L.; Boschetti, G.; et al. Human factors in cobot era: A review of modern production systems features. J. Intell. Manuf. 2023, 34, 85–106. [Google Scholar] [CrossRef]
  5. Zafar, M.H.; Langås, E.F.; Sanfilippo, F. Exploring the synergies between collaborative robotics, digital twins, augmentation, and industry 5.0 for smart manufacturing: A state-of-the-art review. Robot. Comput.-Integr. Manuf. 2024, 89, 102769. [Google Scholar] [CrossRef]
  6. Zhou, Z.; Liu, H.; Dai, Y.; Qin, L. A Tent-Lévy-Based Seagull Optimization Algorithm for the Multi-UAV Collaborative Task Allocation Problem. Appl. Sci. 2024, 14, 5398. [Google Scholar] [CrossRef]
  7. Wu, B.; Wu, X.; Hui, N.; Han, X. Trajectory Planning and Singularity Avoidance Algorithm for Robotic Arm Obstacle Avoidance Based on an Improved Fast Marching Tree. Appl. Sci. 2024, 14, 3241. [Google Scholar] [CrossRef]
  8. Paxton, C.; Hundt, A.; Jonathan, F.; Guerin, K.; Hager, G.D. CoSTAR: Instructing collaborative robots with behavior trees and vision. In Proceedings of the 2017 IEEE International Conference on Robotics and Automation (ICRA), Singapore, 29 May–3 June 2017; pp. 564–571. [Google Scholar] [CrossRef]
  9. ISO/Ts 15066:2016; Robots and Robotic Devices—Collaborative Robots. International Organization for Standardization: Geneva, Switzerland, 2016.
  10. Zhang Sprenger, C.; Corrales Ramón, J.A.; Baier, N.U. RTMN 2.0—An Extension of Robot Task Modeling and Notation (RTMN) Focused on Human—Robot Collaboration. Appl. Sci. 2023, 14, 283. [Google Scholar] [CrossRef]
  11. An, K.; Wu, Z.; Shangguan, Q.; Song, Y.; Xu, X. Robust Learning from Demonstration Based on GANs and Affine Transformation. Appl. Sci. 2024, 14, 2902. [Google Scholar] [CrossRef]
  12. Lin, H.; Quan, P.; Liang, Z.; Wei, D.; Di, S. Enhancing Safety in Automatic Electric Vehicle Charging: A Novel Collision Classification Approach. Appl. Sci. 2024, 14, 1605. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Boschetti, G.; Bottin, M.; Minto, R. AI Technologies for Collaborative and Service Robots. Appl. Sci. 2024, 14, 11811. https://doi.org/10.3390/app142411811

AMA Style

Boschetti G, Bottin M, Minto R. AI Technologies for Collaborative and Service Robots. Applied Sciences. 2024; 14(24):11811. https://doi.org/10.3390/app142411811

Chicago/Turabian Style

Boschetti, Giovanni, Matteo Bottin, and Riccardo Minto. 2024. "AI Technologies for Collaborative and Service Robots" Applied Sciences 14, no. 24: 11811. https://doi.org/10.3390/app142411811

APA Style

Boschetti, G., Bottin, M., & Minto, R. (2024). AI Technologies for Collaborative and Service Robots. Applied Sciences, 14(24), 11811. https://doi.org/10.3390/app142411811

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop