Abstract
Collaborative robotics is one of the key enabling technologies of Industry 4.0. Given the continuous growth of the market, one of the most promising applications of industrial human-robot interaction (HRI) is collaborative assembly. In this context, since prevention from mechanical hazards is justifiably perceived as the primary requirement, cognitive ergonomics and human factors are often overlooked. Cognitive ergonomics should be considered mainly in the design of automated, high-tech, or complex systems. These aspects are strictly related to operators' safety, well-being, and work-related performance. Consequently, roboticists and companies need to consider cognitive ergonomics to develop human-centred and efficient Collaborative Assembly Systems (CASs). This means that interdisciplinary research in the field is needed. Thus, the current proposal focuses on human factors in industrial HRI and, in particular, proposes developing and validating a new cognitive workload (CWL) assessment method based on improvements to the Cognitive Load Assessment for Manufacturing (CLAM) model initially proposed by Thorvald et al. (2017). The new method would ultimately allow limiting the subjective factors when designing CASs and, most importantly, adequately addressing the CWL of operators and how design elements could influence it. Furthermore, a tool for CAS designers will be developed based on the improved CLAM methodology. This practical tool is intended for non-experts in human factors but for experts in assembly systems, who are the primary intended users. Experimental validation of the model and the related tool will be part of the project activities, ultimately providing practical solutions for integrating cognitive requirements in human-centred CASs. The experimental validation will be based on a case study in a realistic context. RESULTS ACHIEVID The ProMentoHR project delivered HERALD (Human-centric Evaluation of Robotics and Assembly cognitive LoaD), a worker-centric method and companion tool for assessing the cognitive workload of operators in collaborative assembly systems. The project was carried out by an interdisciplinary consortium designed to bridge a persistent gap between engineering and the human sciences. The Smart Mini Factory laboratory for Industry 4.0 of the Free University of Bozen-Bolzano (UNIBZ), led by Luca Gualtueri, contributing expertise in manufacturing systems, collaborative robotics, and experimental testing environments, and a team from the Human Factors, Risk and Safety research unit of the Department of Psychology "Renzo Canestrari" of the University of Bologna (UNIBO), led by Marco De Angelis, contributing expertise in human factors, cognitive ergonomics, and work and organisational psychology. This transdisciplinary composition was essential to properly assess the role played by several factors such as task demands, workstation conditions, and the characteristics of human-robot interaction, determinants that are usually addressed in isolation across separate disciplines. The method was built on the CLAM model and extended to capture the cognitive demands that arise specifically from working alongside a robot. Development followed a structured, auditable process that combined a review of the CLAM toolkit, a research-team brainstorming session, and a targeted literature review, resulting in an initial set of 14 candidate factors. These were then refined through a two-round online Delphi study with an interdisciplinary panel of international experts in industrial engineering, robotics, ergonomics, human factors, and human-robot collaboration. Of 37 experts contacted, 15 joined the panel, with 14 taking part in each round. Using structured consensus criteria, expert ratings and written feedback were used to merge, reword, and reduce the initial list into 11 validated factors, organised across three practical domains: the task, the workstation, and the robot/interaction. Each factor was operationalised through clear indicators and a standardised scoring approach, translating expert consensus into a transparent, usable assessment structure. The validated factor-and-indicator structure was then engineered into the HERALD tool, an operational, freely available (Open Access) instrument delivered together with a practitioner-oriented handbook that guides its application step by step. The tool enables automatic calculation of cognitive workload with transparent scoring and adaptable weighting, produces a factor-level profile of a workstation, and offers direct guidance on where to prioritise design interventions. A central design goal was accessibility. HERALD is intended to be used without specialised human-factors expertise by a range of industry professionals, such as Human Factors, Health and Safety experts, product managers, industrial and process designers, and occupational psychologists, to assess operators' cognitive workload. To confirm its usability and practical value, the tool has already been applied by industrial expert end-users in real collaborative assembly settings, whose feedback informed its refinement. In continuity with the project, the first experimental validation sessions have been conducted at the Smart Mini Factory laboratory, across three collaborative assembly workstations with varying configurations and complexity levels. By controlling variables such as work intensity, human-robot communication, robot predictability, and robot adaptability, these sessions are designed to test the tool's sensitivity, validity, and robustness against independent objective and subjective measures. Preliminary analysis is underway, with early indications consistent with expectations, and the full results will be reported in a dedicated scientific publication. The project's outcomes are documented in a scientific article on the method's development and validation, in the HERALD Handbook, and in the Open Access tool, and are shared through the project website (https://site.unibo.it/hfrs/en/projects/promentohr). Beyond the current funding period, work continues to consolidate the experimental evidence and to extend HERALD along the directions identified during the project, including inter-rater reliability, sensitivity to workstation redesign, and refined weighting schemes. By making a transparent, ready-to-use workload assessment method freely available, ProMentoHR aimed at lowering the barrier to integrating cognitive ergonomics into the design and evaluation of human-centred collaborative assembly systems, supporting more standardised, traceable, and worker-centric decisions in manufacturing.
Dettagli del progetto
Responsabile scientifico: Marco De Angelis
Strutture Unibo coinvolte:
Dipartimento di Psicologia "Renzo Canestrari"
Coordinatore:
ALMA MATER STUDIORUM - Università di Bologna(Italy)
Contributo totale Unibo: Euro (EUR) 91.000,00
Durata del progetto in mesi: 24
Data di inizio
28/09/2023
Data di fine:
31/12/2025
