Watch-Bot

Once in a while, even the most meticulous employee can miss something important, and that can lead disastrous results. Now, imagine if technology could help ensure that you don’t have a mishap at work again, and all simply by watching you.

It may cause a number of individuals pause (understandably), but if implemented in positive ways, it could help prevent injury and inefficient work practices by pointing out what you missed—or did wrong—to bring it to your attention and sort it out immediately.

This is a scenario that may soon become a reality with the Watch-Bot, which is a piece of technology that can learn a task by watching humans. And it can be used to maintain, assist. and even supervise quality control by reminding people when they’ve missed a step.

The watch-bot, the brainchild of Ashutosh Saxena adn Chexia Wu from Stanford University and Cornell University, respectively, was created by mounting a depth-sensing camera on a tripod and connecting it to a computer and a laser pointer.

It learns unaided and spots patterns in human movement as it observes.

Assembly Lines
U.S. Navy National Museum of Naval Aviation

In several trials, Watch-Bot was around 60 percent accurate in terms of pointing out missed tasks due to human error using its laser pointer.

Most evident is the watch-bot’s potential for use in assembly lines and manufacturing, but the technology is also being eyed for application for tracking small, mundane daily activities (such as making sure that you don’t forget to bring the keys before leaving the house). It could also be used in safeguarding safety in the household, such as switching the stove or the iron off. Eventually, it can be used to aid elderly people who want to continue to live their lives independently or even be used in homeland security or the military.

The researchers behind the Watch-bot are bent on refining the technology before it is applied to major and vital safety tasks, and determining exactly whether the robot can actually fill a need in the market. 

But the team has take some pretty interesting and promising first steps in their work.


Share This Article