Rethinking Human Ergonomics in Lean Manufacturing and Service Industry: Towards Adaptive Robots with Anticipatory Behaviors

Información del proyecto

Ergo-Lean

Identificador del acuerdo de subvención: 850932

DOI

10.3030/850932

Proyecto cerrado

Fecha de la firma de la CE 21 Octubre 2019

Fecha de inicio 1 Noviembre 2019

Fecha de finalización 31 Octubre 2024

Financiado con arreglo a

• EXCELLENT SCIENCE - European Research Council (ERC)

Coste total

€ 1 488 750,00

Aportación de la UE

€ 1 488 750,00

Coordinado por

FONDAZIONE ISTITUTO ITALIANO DI TECNOLOGIA
Italy

Periodic Reporting for period 4 - Ergo-Lean (Rethinking Human Ergonomics in Lean Manufacturing and Service Industry: Towards Adaptive Robots with Anticipatory Behaviors)

Período documentado: 2024-05-01 hasta 2024-10-31

Resumen del contexto y de los objetivos generales del proyecto

Ergo-Lean aimed to redefine occupational ergonomics by examining complex human-robot-environment interactions and developing predictive methods to anticipate the impact of worker actions. The central objective was to leverage the potential of collaborative robotics to deliver anticipatory behaviors that enhance human psycho-physical well-being. To achieve this, the project focused on five core Scientific Objectives (SOs): SO1: Develop real-time, data-driven risk identification methods for human-environment interactions to improve safety and ergonomic outcomes. SO2: Advance human-robot collaboration (HRC) technologies to mitigate occupational risks through active ergonomic interventions. SO3: Design and implement Ergo-Assistant Interfaces that enhance motion efficiency while reducing fatigue and the risk of injury. SO4: Investigate shared authority models to integrate human intent with robotic assistance, enabling optimal ergonomic conditions. SO5: Validate the adaptability and acceptability of Ergo-Lean systems across diverse work environments and worker profiles.
By achieving these objectives, Ergo-Lean introduced significant cross-disciplinary technical progress in areas such as human dynamic modeling, collaborative robotics, and the use of multimodal feedback to enhance workplace ergonomics. Numerous studies, conducted on hundreds of volunteers, provide concrete evidence of substantial improvements in human factors during (simulated) industrial tasks. These studies also demonstrate increased acceptability and trustworthiness of collaborative robots among novice users.
The direct impact of Ergo-Lean extends beyond reducing work-related musculoskeletal disorders (WMSDs); it also enhances the potential of collaborative robots in industrial applications, fostering safer, more efficient, and ergonomically optimized work environments.

Trabajo realizado desde el comienzo del proyecto hasta el final del período abarcado por el informe y los principales resultados hasta la fecha

A key achievement has been the development of a real-time, personalized human kinodynamics models, enabling precise quantification of motor capacity, ergonomic factors, and metabolic costs. This framework identified overloading joint torques and external contact forces, leading to validated models that improve ergonomic assessments. Using these models, an advanced control framework was developed for collaborative robots, integrating priority-based motion control to optimize human ergonomics in co-manipulation and teleoperation.
The project also introduced novel Ergo-Assistant interfaces, including real-time graphical feedback and vibrotactile guidance, effectively minimizing ergonomic risks. Shared authority models optimized task allocation in HRC, distributing actions based on ergonomic risk and human capabilities, enhancing efficiency while reducing strain. Adaptive HRI frameworks enabled robots to adjust behavior based on cognitive and physical workload using reinforcement learning and optimization techniques. A major breakthrough was the development of a sustainable, biodegradable, solar-powered wearable device for ergonomic feedback. Additionally, a versatile hierarchical control framework was created for various HRI scenarios, eliminating the need for multiple architectures and enabling seamless transitions between interaction modes.
Kinesthetic feedback-driven robotic assistance was another key innovation, where robots provided real-time resistance to non-ergonomic postures, fostering long-term ergonomic habits. User studies, electromyography analyses, and industrial trials confirmed significant reductions in physical and cognitive workload. The project’s impact was widely disseminated through high-impact publications, keynote presentations at leading robotics conferences, and extensive media coverage. The team organized workshops and scientific events, further contributing to human-centric robotic systems. The project's achievements were recognized through prestigious awards, including the IEEE Robotics and Automation Society Early Career Award.

Avances que van más allá del estado de la técnica e impacto potencial esperado (incluida la repercusión socioeconómica y las implicaciones sociales más amplias del proyecto hasta la fecha)

The project adopted an unconventional approach to human ergonomics. Unlike the simplified or kinematics-based models traditionally used to assess ergonomics in industrial settings, Ergo-Lean developed a quantifiable, dynamic model of the human body to reveal real-time body states during interactions with the external environment. This model identified overloading joint torques and musculoskeletal strain, laying the groundwork for next-generation ergonomic assessments.
A key breakthrough was that these models not only enabled continuous monitoring but were also integrated into robotic and automation architectures to actively mitigate risks to human health. While the project primarily focused on physical ergonomic factors, an unexpected outcome was the development of a cognitive ergonomics monitoring system, which combined physiological and behavioral data to provide real-time robotic assistance based on both cognitive and physical workload.
This cognitive system incorporated reinforcement learning and model-predictive optimization, ensuring smooth and adaptive human-robot interactions. Another innovation was the design of a biodegradable, solar-powered wearable feedback system, setting a new benchmark for sustainable ergonomic support technologies.
To enhance functionality, the project introduced a hierarchical control framework that unified various robotic assistance modes, enabling seamless transitions between physical support, vibrotactile guidance, and visual feedback. These advancements were validated through extensive industrial trials, electromyography (EMG)-based workload assessments, and longitudinal user studies, all of which demonstrated the system’s effectiveness in real-world applications.
During the final development phase, the project focused on refining these solutions for large-scale deployment, ensuring smooth integration into both industrial and assistive robotics. Further optimizations to the cognitive adaptation framework enhanced the robot’s autonomous capabilities in mitigating ergonomic risks.
By achieving these outcomes, Ergo-Lean redefined human-centric robotics and established a scalable foundation for future research in ergonomic human-robot collaboration.

picture1.jpg

picture2.png