Assembly lines often face variability in operator performance, causing inefficiencies
that can lead to prolonged throughput times and suboptimal resource utilization. This
paper explores how Digital Twins (DT) integrated within a Decision Support System
(DSS) can address these issues by simulating and re-balancing assembly lines in real-time.
Real-time data is used to update DT predictions including learning effects to generate
alternative solutions in terms of task allocation. Numerical experiments on a simplified
assembly line structure show consistent reductions in cycle time when re-balancing
and adjusting re-balancing intervals.
