https://www.reddit.com/r/maybemaybemaybe/comments/1ja8irf/maybe_maybe_maybe/
From the maybemaybemaybe community on Reddit: maybe maybe maybe
reddit.com
Thanks Melvin for the example. The linked video shows AMRs stuck in an infinite loop due to an obstruction. This is something I started to explore in my work on a 2.5D visualisation of smart warehouses. Each AMR executes a set of rules from a first person perspective. Rules can be used to give priority to one AMR over another. A complementary approach is to introduce random variations in behaviour so that the initial conditions don’t repeat. A more complicated approach introduces episodic memory to mimic how humans remember what they did and try something different. It may also make sense to allow AMRs in close proximity to exchange information on their actions, e.g. “I am slowing down and want to turn right”.
Dave Raggett <dsr@w3.org>
