Much of the past and current research focus on central planning. Just my opinion, but designing simple but robust systems that emulate how simple insects behave is more interesting to me than building expensive and fragilerobots that are supposed to function like humans.