Balancing human-robot mixed-model assembly lines considering ergonomic risk, metabolic energy expenditure, and robotic energy consumption

Abstract In recent decades, automation of mixed-model assembly lines has increased, driven mainly by competitiveness and profitability. However, budget constraints often limit the available investment, sometimes leading to the implementation of semi-automated configurations (including robots and human operators). This paper proposes a lexicographical approach to assembly line balancing that integrates human-centric criteria alongside traditional productivity goals. The methodology involves two initial phases that follow conventional line-balancing practices, achieving takt-time compliance while improving workload balance and reducing idle time. In sequence, Phase 3 uses ergonomic assessments to eliminate high-risk tasks, resulting in an 18.33% improvement in ergonomic risk evaluation and a 6.66% decrease in energy expenditure, demonstrating simultaneous social and performance benefits without requiring additional investment. Moreover, Phase 4 targets energy consumption in welding operations, resulting in a 1.25% reduction (across the total energy consumed), mainly by reallocating high-current spot welds to more efficient guns, which is non-negligible in the context of operational margins and Joule heating requirements. Innovative Mixed-Integer Linear Programming (MILP) approaches were proposed to embed considerations about health and well-being of operators into the line-balancing process, while also reducing energy consumption. Solutions were tested across several production-mix scenarios and ranked using Technique for Order Preference by Similarity (TOPSIS), which weighted social and environmental criteria, revealing that nominal mixes were counterintuitively not the best-performing, underscoring the value of multi-scenario evaluation. The approach handles a large, realistic task set (1413 tasks), advancing applicability to industry and aligning with Industry 5.0’s human-centric ethos.