The Computer Science and Artificial Intelligence Lab (CSAIL) of MIT has released a video of their ongoing work that involves full and fine control of drones. Using just the hand and arm to navigate the drone around through a series of rings, the person can control it.
Instead of using optical or other kinds of gesture recognition, this innovation is not only impressive just because it uses biofeedback to control devices but also because of how specific the controls can be.
The researchers are looking for ways to integrate it into collaborative robotics for potential industrial applications. Drone piloting could benefit from this new technology's real-world use especially when a flock of drones take flight with a pilot provided an aerial view via VR.
This technology can be used to do site surveying for construction or remote equipment inspection of offshore platforms and other infrastructure that people find hard to reach.
Seamless robotic/human interaction
The ultimate goal of the engineers on this technology is to have a seamless robotic/human interaction since that is how humans usually intuit their movements and ability to manipulate their environment most effectively. They believe that the process should be as smooth when controlling and working with drones.
MIT muscle-control system for drones lets a pilot use gestures for accurate and specific navigation https://t.co/D0mEI6xOaG by @etherington pic.twitter.com/YPJcLUr7Gt — TechCrunch (@TechCrunch) April 27, 2020
When people interact with their environment, thinking, and doing happen at the same time. However, when they act through the extension of machines or remote devices, often there is something lost in translation that results in a steep learning curve and requires a lot of training.
The device uses a "plug-and-play gesture control" that relies on muscle and motion sensors. According to their paper published last month, allowing machines to interpret nonverbal commands such as gestures can help make interactions more similar to interactions with another person. But for it to work, it should not require significant sensing infrastructure or per-user setup time.
The device which looks like a bracelet picks up motion through electrical signals and can distinguish between a clenched fist, hand rotation, or forearm movement. The drone was adept at reading desired inputs by 81% out of 1,535 "unstructured gestures."
Cobotics or collaborative robots, is the industry that focuses on building robots that can safely work alongside humans or in close collaboration with robots would benefit from advances that make interaction more natural, instinctive, and ultimately, safe.
This research by MIT in this area could result in future industrial robotics products that will require less training and programming to operate the device at scale.
Cobots assist humans by augmenting their capabilities in terms of effort. This will allow humans to manipulate parts that are hot, heavy, bulky, or too small for precision handling. Since humans will work close with machines, they remain necessary for detailed observation skills, and deal with complex parts of the job.
Contrary to popular belief, although the factory of the future will feature more automation, people who are the source of innovation will remain relative. The staff who performs operations and protective maintenance are one of the foundations of the future factory.