In Part 1 of these lectures, the INFORM Lab testbed was described with respect to its general architecture, OODA nodes, multi-layer plug-and-play architecture, and general multi-agent framework based on John Boyd's OODA loop. Part 2 demonstrates some promising results from a non-cooperative scenario involving complex behaviour between multiple ships trying to hide their motives, as would be the case in a piracy scenario. The flexibility of the testbed is also demonstrated.