You might find it interesting that the next pair of Gap jeans you buy may be delivered to you with the help of a futuristic robot.

A robotics startup company, Kindred, has signed up with Gap Inc. to become their trial customer in one of its warehouses in Tennessee. The three-year-old Kindred made a big win bagging Gap as their first client especially after scoring $28 million in funding from a Chinese tech giant recently.

The company’s new customer together with its new investment continues to highlight the growing trend of companies such as Target and Amazon that use robots to move their warehouse inventory and keep track of their products on their shelves. Many robot analysts cited the $775 million purchase of Amazon on Kiva Systems back in 2012 as a turning point for retailers and other organizations in the adoption of robots.

The robots from Kindred, which works by sorting items shuttled from a conveyor belt, look similar to those found in old arcade games which involve the use of claws attached to cranes intended to pick up plastic trinkets. Kindred’s robots can pick up packages that contain things such as sunglasses, t-shirts, and pants which are then dropped off inside a glass-paneled enclosure in the robot. They would then swivel around and put these items in their designated drop-off zones.

The co-founder of Kindred and their product chief, George Babu, mentioned that Gap has been testing their robot for quite a while now and it is now in the works to process real-time orders. It does so by helping humans separate things such as t-shirts and jeans for delivery.

At this time, the staff from Kindred in Canada are helping Gap train their robots remotely for the automated work being done in their Tennessee warehouse. Babu also mentioned that the underlying software which powers their robotic arms are built by the Japanese robotics company Fanuc. If any problems occur on the assembly line, the staff from Kindred can immediately notify the appropriate operator in the warehouse about it.

With the use of custom software, Kindred’s staff manipulate these robots as if they were playing a video game. Every time a Kindred employee moves a robotic arm to pick something up, the software would record this data so that the robot would be able to learn how to do this without human intervention.

Unlike the robots that we have seen pop up in the news previously, the robots from Kindred are made to improve over time with the help of a subset of artificial intelligence known as deep learning. Although companies such as Facebook and Google have been using deep learning to sift through data, the robots from Kindred are using it to help remember motions that are needed for the job.

Aside from that, these robots also use AI technology called reinforcement learning. This helps the machines figure out the best way to pick up and move items to their drop-off spots.