When we studied robotics and artificial intelligence in class, we watched videos of the DARPA Grand Challenge – a race for autonomous vehicles through the Nevada desert, and the DARPA Urban Challenge – where autonomous vehicles had to navigate streets, avoiding traffic and obeying the rules of the road. Stanley, the car entered by Stanford University, won the Grand Challenge and also completed the Urban Challenge.
Now DARPA are about to announce a new challenge – for designers of robots which can operate in disasters zones. In class we have seen examples of many robots used in dangerous situations: a snake-like robot for searching collapsed buildings, bomb disposal robots, robots used to investigate the Fukushima nuclear power plant, and many more.
However, for this challenge DAPRA want something completely different: the robot entries must be humanoid, able to interact with and use human tools and equipment. DARPA says the winning entries must be able to:
- Grasp and use human tools, rather than having attached tools as many robots do. This includes using a key to open a lock door.
- Enter a standard vehicle, drive it, and then dismount.
- Climb a ladder
- Locate a leaking pipe
- Seal the leak
The brief says the robots must be semi-autonomous, with just “supervisory tele-operation” from human operators. The challenge will begin in October 2012.
- List four specific examples where robots like this could be deployed.
- What do you think are the benefits of using robots in situations like this?
- Why do many current designers avoid creating humanoid robots?
- Why do you think DARPA want humanoid robots for this challenge?
- What is the difference between a humanoid robot and an android?
- What input devices and techniques could such a robot use to sense its environment?
- What input devices and techniques could such a robot use to complete its job (this is a different question from the one above!)