Polite robots running various errands could someday join the people who stream through downtown areas, striding or rolling along without jostling people or getting in their way, thanks to research underway at MIT.
MIT engineers have developed a small robotic vehicle that can seamlessly move with people in public spaces. The technology could eventually lead to robots that perform a variety of delivery and transportation tasks.
The autonomous robot uses “socially aware navigation” software to keep pace with foot traffic and observe the “general codes of pedestrian conduct,” the Massachusetts Institute of Technology News Office said.
The robot, described as a “knee-high kiosk on wheels,” completed testing at MIT’s Ray and Maria Stata Center that demonstrated its capability to avoid collisions with people while keeping up with average walking speeds.
“Socially aware navigation is a central capability for mobile robots operating in environments that require frequent interactions with pedestrians,” said MIT graduate student and researcher Yu Fan Chen. “For instance, small robots could operate on sidewalks for package and food delivery. Similarly, personal mobility devices could transport people in large, crowded spaces, such as shopping malls, airports, and hospitals.”
Chen is the lead author of a research paper on the robot’s programming that will be presented at an Institute of Electrical and Electronics Engineers conference in September.
Fellow MIT graduate student Michael Everett, former postdoctoral researcher Miao Liu, and MIT professor of aeronautics and astronautics Jonathan How co-wrote the study.
MIT said the researchers dealt with three of four main autonomous movement challenges using fairly standard approaches, while developing a new method for the fourth.
Open-source mapping algorithms were used to locate the robot in the environment, “off-the-shelf” sensors were used for its perception, and controls similar to those employed in autonomous cars helped direct the robot.
The fourth problem, engineering the robot to successfully navigate around changing pedestrian conditions, proved the most difficult.
“The part of the field that we thought we needed to innovate on was motion planning,” Everett said. “Once you figure out where you are in the world, and know how to follow trajectories, which trajectories should you be following?”
People rarely stick to straight, predetermined paths, tending to weave and wander, veering off to greet a friend or grab a coffee. That’s a challenge for robots that had been unsolved.
Currently, “people don’t find [robots] to fit into the socially accepted rules, like giving people enough space or driving at acceptable speeds, and they get more in the way than they help,” Everett said.
To solve the motion planning problem, the researchers used a reinforcement learning approach, in which the robot was trained through simulations showing the robot how to take certain paths while following social norms.
“We want it to be traveling naturally among people and not be intrusive,” Everett said.
This method keeps the robot readjusting its path every one-tenth of a second, allowing it to roll at a typical walking speed of 1.2 meters per second without colliding with pedestrians or having to stop and reprogram its route.
Everett said the robot “trained” in the Stata Center’s hallways, rolling alongside students going about their everyday activities.
“One time there was even a tour group, and it perfectly avoided them,” Everett said.
The research team eventually hopes to put the robot in even more crowded pedestrian environments, training it in how to deal with larger groups.
“There may be a social rule of, ‘Don’t move through people, don’t split people up, treat them as one mass,’ ” Everett said. “That’s something we’re looking at in the future.”
Ben Thompson can be reached at email@example.com.