Christopher W. Geib and Robert P. Goldman
Team-level intent recognition presents new and interesting challenges for the task tracking research and violates some of the implicit and explicit assumptions behind much existing research. While our work hasn’t focused specifically on the problem of inferring team intent, we have looked at environments that have many of the same properties. In this extended abstract we briefly discuss one such environment and one of the central problems it presents. We then briefly sketch a plan recognition algorithm we have been working on that provides some leverage on this problem.