Albeit agent-orientation became a well-established course in artificial intelligence at the engineering stage its effectiveness is rather poor for affordable agent-based systems. Despite the increasing number of biologically inspired models, the newer paradigms are in a syncretic stage. Thus, although the fidelity towards the biological model is sometimes quite low, inter-paradigmatic synergy is not manifest enough. The paper aims at: a) boosting such synergy at two levels (ant-like entities and symbolic processing); b) validating its path by testing it on a relevant problem in the field of operational research; c) proposing mechanisms with synergistic potential. Specific mechanisms are designed to graft symbolic components onto the sub-symbolic foundation (the filtered biological model), are tailored lo manufacturing control, and are tested with usual benchmarks on an experimental model. The paper concludes that combining stigmergic coordination with symbolic processing components has significant synergistic potential. The most useful mechanism proved to be '·user-driven heuristics".
Stigmergic coordination (SC), Agent-Based Systems, Synergy, Manufacturing Control (MC), Travelling Salesperson Problem (TSP), User-Driven Heuristics (UDH)