I want to write several posts about autonomy for terminator-style robots.

1. We’re so far from real decision making - the fall-flail vs fall-crawl dichotomy and what it’d take.

2. Tactics 1/N: Implicit coordination

   • There’s something remarkable about to happen in planning, esp for multi robot systems
   • First we should talk about strategy of conflict.
   • Signalling / consensus finding games without communication
   • key factor is that you and your adversary must find a consensus. This is a supserset of the cooerative versions
   • Generally, this can be accomplished many ways in traditional planning
   • But the leap from map to decision is so tenuous
   • (here’s a implementation of the gamse to talk about)

3. Tactics 2/N: Recovery and planned communication

   • I know that he would need to know this, and I would expect him to come tell me so I will go tell him.
   • I can expect to be expected to anticipation the needs of others

4. Strategy 1/N: Surveying the situation

Key addition: Decision models