¹National Research Centre for the Working Environment, Copenhagen, Denmark

Text

Introduction: Warehouse workers perform frequent lifting tasks guided by real-time algorithmic instructions via headsets. Our research group found that this management approach distributes lifting work unevenly, increasing low-back pain risk among those lifting the heaviest loads. This study develops an ‘Ergorithm’ for warehouse management systems (WMS) that distributes lifting tasks more evenly while also integrating employee influence without compromising productivity.

Methods: We collect logistics data about staffing, orders, goods’ weights, volumes, handling times, locations etc. An ErgoScore from 0.0-1.0 is created where levels <0.5 represent e.g. pallets with lighter goods located across longer distance, while levels >0.5 represent pallets with heavier goods located within shorter distance – while maintaining productivity (key performance indicators (KPI)). The ErgoScore keeps a 0.5 average by alternating pallets <0.5 and >0.5. Simulations test the feasibility of the Ergorithm in different situations (fewer staff, busier days etc.). We involved warehouse managers and workers in focus group interviews to explore the feasibility of the Ergorithm, and to gain a deeper understanding of the level and possibility of influence at work, definitions of pallets, use and understanding of WMS etc.

Results: Currently, the project contains preliminary results. More results are available and presented at the conference. Daily ErgoScore means and standard deviations (SD) across workers are smaller and closer to 0.5 with the Ergorithm compared with current algorithms, and SD of daily lifting load means of workers (kg) are smaller with the Ergorithm. Thus, the Ergorithm results in more equally distributed lifting loads across workers without compromising productivity. Simulations confirm theoretical feasibility. Workers view the Ergorithm highly positively, but express concern about maintaining KPI, which remains a large focus for the workers. Discrepancies emerged between workers and researchers’ definitions of good and bad pallets, centred on KPI vs. ergonomic work demands.

Discussion: Employing the Ergorithm in WMS possess large potentials and may improve the working environment by increasing variation in work intensity (lifting loads) during the workday, allowing restitution periods without compromising productivity. More equally distributed lifting loads reduces very high daily lifting loads – a significant risk factor for low-back pain. Considering the workers’ knowledge and perceptions is important when developing the Ergorithm. Integrating employee influence in the Ergorithm may concurrently improve the psychosocial working environment. E.g. including individual preferences, competencies, resources etc.

Conclusion: Algorithmic management may improve the working environment among warehouse workers. The Ergorithm may increase variation in work intensity during the workday and distribute lifting loads more equally. Warehouse workers are positive about the Ergorithm, contingent on KPI maintenance. The Ergorithm may be a sustainable solution for creating a healthy working environment through algorithmic management and preventing musculoskeletal disorders.