Human communication is a very complex process, and in addition to words, many other factors come into play, such as tone, non-verbal language, and even pheromones. Similarly, the human brain can predict another individual’s movements by assessing context and past experiences. For example, if you are driving and reach a pedestrian crossing where someone is waiting, you can predict their movements and how long it will take them to cross the road. We may also unknowingly anticipate non-linear movements, i.e. acceleration or deceleration. And if you’re a juggler, stand in the middle and get a tip.The problem is robot They are not good at assessing context and predicting events. This is what BMW engineers found in simulations. Robot-human interaction.
first experiment with this technology project When done in 2018 in collaboration with MIT researchers, they created a replica of the factory. There, robots had to move through rails carrying equipment while workers crossed from one side to the other. Once the robot sensed human movement, it remained completely still until the operator crossed it. The software could not predict how much time was spent traversing and how long it would take to return.Therefore, they decided to develop a new Algorithms applied to robotics This can significantly reduce productivity in large manufacturing environments.Result is new algorithm Because the human partial trajectory can be predicted in real time, robot Measure the time you actually have available so you can move without risking it.
a matter of time
new Artificial intelligence The system considers several factors. Not only the distance traveled, but also the timing can be evaluated. For example, if someone just started moving, it probably takes longer to get back to the starting position. You must first reach your destination, do what you have planned, and then return. Similarly, if they cross in one direction, they are less likely to return quickly. Instead of following a linear movement pattern, the software pulls information from a database of thousands of different movements. This allows them to learn like humans do. According to Julie Shah, associate professor of aerospace science at MIT, this is his one of multiple approaches to better understand human behavior with robots.Use of these techniques robot technology It can also apply to many other everyday situations, such as automotive traffic situations where the unexpected is the norm.
sauce: When
