If you’ve ever had a robot in your home (most likely a pool cleaning robot or a robotic vacuum cleaner), then you’ll be familiar with how basic their movements are – usually in deliberate straight lines without much knowledge of what’s in front of them (until they bump into a wall or object, rotate and go off on another straight line). These robots aren’t what you’d call collision-free.
Well, this is mostly because home-ready robots are super basic in that not much thought is put into their directions of travel, because it’s not that important for what they’re doing at the moment.
This is, however, not going to be the case for robots forever, as a new algorithm developed by researchers at the University of South Australia is set to help robots avoid obstacles in their path.
“There are two types of path planning strategies for mobile robots, depending on whether they are being used in fixed environments or where they are encountering moving obstacles, such as humans or machines,” says Dr. Habib Habibullah, a mechatronics engineering lecturer at UniSA.
“The first is fairly easily to program but the second is more challenging.”
The collision-free robot computer model built by Habibullah is set to teach robots how to avoid obstacles in their path and effectively map out the most optimal route to reach their destination.
In his paper published in the Journal of Field Robotics, Habibullah describes how he and his colleagues created a collision-free Turtlebot, with the ability to adjust speed and angles to respond to emerging collisions. This algorithm is sophisticated enough to stop the robot, take a safe turn and even reverse.
“Our proposed method sometimes took a longer path, but it was faster and safer, avoiding all collisions.” Habibullah added.
The model created by Habibullah was tested against other commonly available algorithms, in particular the Dynamic Window Approach algorithm and the Artificial Potential Field algorithm. In testing, the new algorithm was found to be much more effective in adjusting to new obstacles across nine different scenarios.
So, what are the applications of such collision-free robots? Of course they could be used in more domestic settings, like in robotic vacuum cleaners, but Habibullah reckons the practical applications of this AI are far greater, believing it could be used in fruit picking, in industrial warehouses, in packing, in lawn mowing and also in delivering food to tables in restaurants.
Just another win for our future robot overlords.