Human/Robot Cross-Training Touted in MIT Study

"This is the first evidence that human-robot teamwork is improved when a human and robot train together by switching roles, in a manner similar to effective human team training practices," Ph.D. student Stefanos Nikolaidis said.

A paper to be presented at the International Conference on Human-Robot Interaction in Tokyo next month reports cross-training is an effective team-building tool when human beings and robots are working together, MIT News correspondent Helen Knight reports. The research was done by Julie Shah, an assistant professor of aeronautics and astronautics at MIT and head of the Interactive Robotics Group in the Computer Science and Artificial Intelligence Laboratory, and Ph.D. student Stefanos Nikolaidis.

"People aren't robots, they don't do things the same way every single time," Shah told Knight. "And so there is a mismatch between the way we program robots to perform tasks in exactly the same way each time and what we need them to do if they are going to work in concert with people."

They wanted to find out whether cross-training, an effective techniques for training people, also can work in mixed teams of humans and robots. This entailed creating a new algorithm so the robots could learn from their role-swapping experiences. "So they modified existing reinforcement-learning algorithms to allow the robots to take in not only information from positive and negative rewards, but also information gained through demonstration. In this way, by watching their human counterparts switch roles to carry out their work, the robots were able to learn how the humans wanted them to perform the same task," Knight reported. Each mixed team then performed a simulated task in a virtual environment, with half of the teams using the conventional interactive reward approach and half using cross-training and switching roles halfway through the session. The teams then did the task in the real world, but this time sticking to their own designated roles.

"Shah and Nikolaidis found that the period in which human and robot were working at the same time — known as concurrent motion — increased by 71 percent in teams that had taken part in cross-training, compared to the interactive reward teams. They also found that the amount of time the humans spent doing nothing — while waiting for the robot to complete a stage of the task, for example — decreased by 41 percent. What's more, when the pair studied the robots themselves, they found that the learning algorithms recorded a much lower level of uncertainty about what their human teammate was likely to do next — a measure known as the entropy level — if they had been through cross-training," according to her Feb. 11 report. "Finally, when responding to a questionnaire after the experiment, human participants in cross-training were far more likely to say the robot had carried out the task according to their preferences than those in the reward-only group, and reported greater levels of trust in their robotic teammate."

"This is the first evidence that human-robot teamwork is improved when a human and robot train together by switching roles, in a manner similar to effective human team training practices," Nikolaidis said.

Product Showcase

  • AirChek Connect Sampling Pump

    Stay connected to your sampling with the SKC AirChek® Connect Sampling Pump! With its Bluetooth connection to PC and mobile devices, you can monitor AirChek Connect pump operation without disrupting workflow. SKC designed AirChek Connect specifically for all OEHS professionals to ensure accurate, reliable flows from 5 to 5000 ml/min and extreme ease of use. AirChek Connect offers easy touch screen operation and flexibility. It is quality built to serve you and the workers you protect. Ask about special pricing and a demo at AIHA Connect Booth 1003. Read More

  • Magid® D-ROC® GPD412 21G Ultra-Thin Polyurethane Palm Coated Work Gloves

    Magid’s 21G line is more than just a 21-gauge glove, it’s a revolutionary knitting technology paired with an advanced selection of innovative fibers to create the ultimate in lightweight cut protection. The latest offering in our 21G line provides ANSI A4 cut resistance with unparalleled dexterity and extreme comfort that no other 21-gauge glove on the market can offer! Read More

  • SwabTek® Cannabis Test Kit

    The SwabTek® Cannabis Test Kit is a single-use spot test designed for use in screening for cannabis compounds in any sample type or on any surface. The test is capable of identifying the presumed presence of cannabinoids in very small quantities, with a level of detection as little as 6 μg in mass. Learn more about the SwabTek® Cannabis Test Kit and the rest of SwabTek surface drug testing solutions through the webinar titled "Everything You Want To Know About Surface Testing" Read More

Featured

Artificial Intelligence