Robots for manufacturing. Generating a new family of robots that have a
fluent interaction with humans. It will be easily programmable. Some would
argue that Baxter provides this functionality. Unfortunately Baxter is
slow, there is no abstraction in programming and it does not have the
required accuracy. There is plenty of room to make this a real success. We
in the field of robotics need more such successes.
The most pressing problem to consider is the design of more robust, flexible, and
affordable grippers for robots. This is a major limitation for more widespread
use of robots in many different applications. Parallel pneumatic grippers are
affordable, but the more dextrous systems are much too expensive.