MechSE assistant professor João Ramos and Computer Science associate professor Kris Hauser were recently awarded a three-year NSF grant to investigate the teleoperation of wheeled humanoid robots.
“The idea is to study how robots can help people do their jobs better,” Ramos said. “We will try to develop a robot that enables remote manual labor.”
Funded through the National Robotics Initiative, the research team will investigate the development of a human-like robot, complete with manipulatable arms, that locomotes on wheels and can perform physically demanding tasks like pushing or lifting heavy payloads. The robot will be controlled remotely using teleoperation, meaning that the operator will use full-body haptics to communicate input and receive multisensory feedback.
“An especially innovative part of this research is that we are studying how full-body haptic feedback lets operators feel the weight and impact of their movements, making the experience more immersive,” Hauser said.
The researchers have proposed a bilateral teleoperation framework in which the operator can direct the robot’s locom
otion by leaning forward and send actionable input to its arms using their own upper body gestures. The robot will be designed to address Dynamic Mobile Manipulations (DMM), or physical tasks that require a combination of forceful manipulation and agile locomotion.
“The idea is that if there’s a disaster, you send the robot in and it can be controlled by an operator from a safe position,” Ramos said.
Ramos brings his expertise in robot design to the project, while Hauser contributes expertise in robot motion planning controls and software implementation. “I think we make a really powerful team because we complement each other’s expertise,” said Ramos, who will facilitate development of the prototype and controls while Hauser works to develop safety features for shared autonomy.
The pair have three aims for the three-year study: to implement a whole-body haptic device and humanoid robot system; to study whole-body teleoperation strategies for DMM, including how to map human-to-robot motion and force feedback; and to improve DMM safety by creating predictive control schemes that follow the operator’s commands while preventing input that would damage the robot.
“A lot of what we’ll do in these first three years is to ask the fundamental question, ‘Can we do these very complex full-body tasks via teleoperation?’” Ramos said. “And if we find out that we can, I think there are a lot of really interesting research projects that can be derived [from the work].”
At the end of the three-year period, the robot prototype and full-body haptic system will be tested using an obstacle course that stimulates authentic DMM. “The operator will control the robot in a simulation of a search and rescue scenario that involves pushing debris, breaking down obstacles, and dragging mannequins to safety,” Hauser said.
While both already have graduate students involved in the work, they welcome contributions from undergraduate research assistants, who could develop subsystems for the robot as well as design and build the obstacle course. Interested students should contact João Ramos at jlramos@illinois.edu.