My initial reaction is two-fold. First, simply because no one on the ground is talking about autonomy now does not devalue this discussion in any way. As I pointed out in one of my posts, the reality is that research and development are pushing the autonomy envelope, so in fact, now is exactly when these questions should be brought up. Also, at the risk of sounding dismissive, which is certainly not my intention; however, given that most of us are academics with no experience in the field these macro consideration are precisely where we can make out largest contribution to the discussion.

As for Will's second point, on the application to the battle space, it seems to me that robotics are most applicable to the intelligence and reconnaissance area. As nanotechnology continues to reduce the physical size of these machine, their ubiquitous extension into all ISR seems inevitable; though again, this push must be predicated on a serious discussion of the legal and ethical boundaries of development and application of these machines.

Wow. I didn't know anyone would take my comments that seriously. I like what you guys are doing, but, as originally intended, I would like to inject a note of caution, from the guys actually using these things TODAY.

What the users want is better and more reliable control. They don't want ROBOTs thinking or acting by themselves. The one area I have seen where autonomy is valuable to an extent is logistics, but that is set in a "follow me" me type function.

To put this in perspective, some 80-90% of the cost of one of the current operational ground crawling robots, with a combat record, is in the communications package, it carries. Not sensors, not chassis, or anything else. The effort is going into ensuring control.