Hey Robot, let's swap jobs! // EIT Digital

Hey Robot, let's swap jobs!

Illustration of human-aware motion planning. The left panel depicts a shared workspace in which a human and robot are placing and sealing screws, respectively. The right panel depicts both the standard, shortest-path motion (dashed arrow) and a human-aware motion (solid arrow) that the robot could take given the expected human workspace occupancy, represented by the cylinder. Credit: Lasota, P. A., and J. A. Shah/Human Factors: The Journal of the Human Factors and Ergonomics Society

In Industry 4.0 we will likely see a seamless cooperation of workers and machines (robots of some sort). This is an area we are looking at within the CyberPhysical Systems activities at EIT ICT Labs.

Here is a news from MIT that nicely fits into the path towards Industry 4.0. How to improve the cooperation between workers and robots as part of a project funded by NSF.

It is trickier than it might seem at first glance. Today manufacturing involves plenty of robots and usually there is a cordoned area to make sure robots cannot harm people as they do their job. The problem is that robots today are very focussed on what they have to do (placing components on a board, soldering, painting) and they do not have a perception of what is going on around them (like a person coming within their space of operation). Also, robots can do amazing things with high reliability and precision but their operation needs to be coded in their brain (the computer operating them). Changing this coding is not trivial.

Several studies have taken place to simplify the coding of operation, see Baxter, that now in some cases can be accomplished by "showing" the robot what it is supposed to do. 

At MIT they are studying the dynamics of cooperative working in teams, like military teams, cockpit teams, surgery teams, to learn how to create those dynamics in mixed robots and human teams. This first requires a sort of awareness from the robot side that a human is around, and part of the team. This involves ambient sensors that can ring the information to the robot and an artificial intelligence program that can "understand" what is going on.

Next they are trying to create an awareness, both in the workers and in the robots of what the other participants in the team are doing. This seems to work pretty well in human teams. You take up your team mate job and she takes up your. By doing what you will be doing she gets a much better understanding of your point of view and will be able to cooperate much better adapting her behaviour to your needs and viceversa.

This is what MIT researchers are trying to do. Have workers doing the robot job (may be in at a slower pace and with lower reliability and precision) and having the robot imitate the job that a worker will be doing creating a learning experience that will help the robot in being aware of what the persons around it might be doing.

I find this a fascinating evolution a much more "intimate" relation between us and the robots. Asimov, here we come!

Author - Roberto Saracco

© 2010-2018 EIT Digital IVZW. All rights reserved. Legal notice. Privacy Policy.