A team of researchers at the Robotics Institute of Carnegie Mellon University (CMU) has designed an innovative suite of robotic systems and planners that empower robots to navigate and map unknown and potentially hazardous environments more swiftly and accurately. The Autonomous Exploration Research Team's creations enable robots to explore autonomously, charting their path and creating a map without any human intervention.
The research group from CMU has amalgamated a 3D scanning lidar sensor, forward-looking camera, and inertial measurement unit sensors with an exploration algorithm. This combination allows the robot to establish its current location, previous routes, and future destinations. These sensors can be affixed to almost any robotic platform. Currently, the team is utilizing a motorized wheelchair and drones for most of its testing.
According to Ji Zhang, a systems scientist at the Robotics Institute, the technology can be deployed in any environment, such as a department store or a residential building post-disaster.
"It builds the map in real-time, and while it explores, it figures out where it wants to go next. You can see everything on the map. You don’t even have to step into the space. Just let the robots explore and map the environment," Zhang said.
The system offers three different exploration modes. In the first mode, a human can direct the robot's movement and direction while autonomous systems prevent it from colliding with walls, ceilings, or other objects. In the second mode, a human can choose a point on a map, and the robot will navigate to that location. In the final mode, the robot embarks on an autonomous exploration of the entire area to generate a map.
Over the past three years, the CMU researchers have been focusing on exploration systems like this one. The system has already mapped numerous locations, including several underground mines, a parking garage, the Cohon University Center, and various indoor and outdoor spots on the CMU campus.
Compared to previous robotic navigation and mapping approaches, this system is more efficient. It can create comprehensive maps while halving the operational time. Moreover, it can function effectively in low-light and dangerous conditions where communication is unreliable, such as caves, tunnels, and abandoned structures.
The team's latest work was recently published online in Science Robotics, under the title "Representation Granularity Enables Time-Efficient Autonomous Exploration in Large, Complex Worlds."