Our current crop of military robots includes some of the familiar small, tank-like unmanned ground vehicles (UGVs) that move on treads or wheels. It also includes more specialized types like large autonomous vehicles that can drive themselves and carry big loads, along with remote-controlled versions of earth-moving machines.
Other military robots have wings and can be launched by hand or by tube. Larger versions have vertical takeoff and landing (VTOL) abilities and can be sent to and from otherwise inaccessible locations. Most of these robots carry a wide range of audio, video, sensor, and communications abilities. One being developed by the US Navy carries its own fuel cell.
Click the image below for a slideshow 10 of the most intimidating robots employed by the US military.

(Source: Robotex)
This article was originally published on Design News and republished on EBN's sister publication EE Times.
If, by 'mess with', you mean play with the controller…count me in! If, on the other hand, you mean become the object of the robot's interest….no thanks!
 These are some amazing machines, and they are only a subset of tactical robots that are available,
Â
I agree. I wouldn't want to be the object of the robot's affection either.
@pocharle – The robot's affection wasn't really what I was thinking – but I wouldn't want that either!
These robots would be useful in sending to disaster areas where humans cannot go or its very costly to give away human life. Like search of missing airlinesetc.
@sp: Good suggestion but in disaster areas you do need to take quick decisions, so by sending robots will not be helpful to tackle the situations.Â
I know what you mean. By “affection”, I meant target. I don't think there's algorithms good enough to simulate real emotion.
How do you mean by taking quick decisions? I think robots would be more useful virtually for everything in years to come. Simply for efficiency reasons. Hardly could common errors often caused by human happen in case of robots, unless failure due to hardware or software.
>>I don't think there's algorithms good enough to simulate real emotion<<
Getting that may not be too difficult with the level of advancement World has reached today, especially in technology space
The closest I can see now is something like Watson, but is that truly something that will be used everywhere? Unless the cost of use will be significantly less, then I think it can only be used for large organizations.