Cornell University’s Personal Robotics Lab, which develops software for building high-end robots, is currently focused on creating robots which can effectively perform various daily activities including cleaning up an untidy room, arranging a bookshelf and packing and unpacking a dishwasher, without the interference of a human.
Ashutosh Saxena, Assistant Professor of Computer Science, Cornell University, heads Cornell’s Personal Robotics Laboratory. According to him, in another five to ten years, people will start procuring assistive robots, either at a minimum price or at the same price as that of a car.
One of the major tasks faced by them is including the learning ability to these robots in doubtful environments. For instance, to make the robot to grasp an object it had never come across or to make it to move across a room it had never seen.
Saxena and his students are involved in improvising their codes for making their robots to perform the desired tasks. One of their projects is to develop a robotic arm with a holder. The robot utilizes a camera to assess the object and then it adopts the best possible method to grip it. It was reported that in due course, this technology will be incorporated into the fully developed dishwasher-loading robot.
Another research platform is to create a wandering robot with a camera to check a disorderly room thoroughly for finding an object like a shoe. Saxena said that the searching can be made simple by recognizing the three-dimensional structure of the room.
The Robotics lab is developing learning algorithms for quick integration of numerous observation algorithms into a more consistent one. All these projects were presented by his students Congcong Li and Adarsh Kowdle at the European Conference on Computer Vision, which was held in Greece from September 5 to 11, 2010.