Assignment and sequencing of parts to autonomous workstations

We present an optimization-based coordination protocol among autonomous workstations in a multiprocessor stage devoted to painting of the shutters in a furniture production process. The coordination aims to maximize the number of parallel operations executable at each machine cycle, while fulfilling constraints on the unique-copy tools. The mechanism is derived by a distributed implementation of a bipartite matching algorithm. The resulting procedure is shown to be compatible with the several autonomous decisions characterizing the process.