Imagine a future where robots are taking your Starbucks order, bagging your groceries and even serve as your personal chef! Ashutosh Saxena is leading a new research project in his Robot Learning Lab at Cornell University where he teaches robots how to infer actions based on complex, human comments. This new technology will allow individuals who aren’t computer programmers to casually communicate with robots. Instead of giving robots detailed step-by-step instructions the team is teaching them to understand commands in natural language from various speakers, account for missing information and adapt to the environment at hand. With this technology, you can say, “Make me a bowl of ramen” or “Clean up my mess,” and the robot will perform the task. Equipped with 3-D cameras, the robots use Saxena’s computer vision software to scan its environment and identify objects. They are able to associate objects with their capabilities. For example, a pan can be poured into or poured from and stoves can heat things. They can also pick a number of objects to perform the same task. If you ask a robot to “heat water” it will be able to use the stove or microwave depending on what is available. The robots are still not 100 percent accurate, but do perform the right steps about 64 percent of the time even when the commands change and environment is reorganized. “There is still room for improvement,” said the researchers. You can teach a simulated robot to perform a kitchen task at the “Tell me Dave” website, where your input there will become part of a crowdsourced library of instruction for the Cornell robots. “With crowdsourcing at such a scale, robots will learn at a much faster rate,” said Saxena. Saxena and graduate students, Dipendra K. Misra and Jaeyong Sung, will describe their methods at the Robotics: Science and Systems conference at the University of California, Berkeley, July 12-16.
0 Comments
Leave a Reply. |
FortyThree is a new breed of public relations that works with bloggers, social media and print, online and broadcast journalists to let you join the conversation with your customers... Archives
February 2016
Categories
All
|