Mobility as Living Systems

As part of considerations into how to develop logistics services to transport a wide range of goods, we pursued joint research to explore how living systems could be used to impart added value to Denso’s services.

We believe that services used to transport goods between individuals and companies very closely resemble the way in which life systems develop. Vehicle congestion (convergence) is a phenomenon of self-organization that occurs in logistics. A wide range of research has been performed on this domain, using vehicle speed and density as parameters. In addition to vehicle speed and density used in traditional models of congestion, we also built a model in which a vehicle collects an item from an agent and delivers it to a specific destination or agent. We then observed, given the number of agents and their movements, and the distribution of destinations, what logistics patterns existed in the system. Notably, we were able to visualize the way in which vehicles obtained goods from agents and moved them to other locations.

This visualization is reminiscent of ant colony optimization (ACO), famously used in the field of ALife, but the main difference is that while ACO synchronizes the movement of all agents in order to enable more rapid transport of food, the congestion model uses synchronization as a cause of convergence. ACO seeks to optimize synchronization and consider the resulting feedback. In the congestion model, while synchronization is counterintuitive to transport, we sought to consider what mechanisms might be beneficial.

We turned to asynchronicity of the system, with the research focusing on cooperative phenomena, something new to this domain of research. Unlike a computer architecture which does not function without a global clock, the asynchronicity seen here has one benefit, namely that it can be spatio-temporally scaled. The notion of indefinite scalability in the construction of ALife systems will likely continue to be debated at length going forward.