28-May-2021 | Market Research Store
Over the years, robots have been equipped with a range of features such as computer vision modules, artificial touch sensors and some advanced humanoids have even been altered with smell sensors. Researchers are often keen on integrating robots with human-like senses. The researchers at Massachusetts Institute of Technology have developed a robot that implemented radio technology in order to detect objects and navigate its way through. The robot, or GF-Grasp, contains a combination of powerful sensing and traditional computer vision in order to locate and grab objects that are in its view of line.
The following advancements are being made in mind for a functioning e-commerce warehouse. As e-commerce continue to flourish, the extreme logistics behind every order placed will require more care and manpower in order to be implemented completely. Warehouses now implement an automatic system for easy functionality working alongside its human counterpart. However, robots struggle in dealing with perception to locate and grab an object in crowded environment as computer vision struggles to deal with working in 3D space in its best version. As computer vision fails to overcome such challenges, radio waves were integrated into the program and research.
The radio waves provide infinite information about the location and identity of the object being in play. Radio frequency works on a reader and a tag – a small computer chip that is used as the target for identification. The reader emits an RF signal which can be interpreted by onboard computers for an approximate location. RF Grasp implements radio technology using a combination of computer vision and an RF reader in order to locate and grab objects even if they are fully blocked from the initial view. It further consists of a modified take on the traditional robotic arm which sits on the robots wrist. The camera and the RF reader constantly create a virtual environment of RF Grasp’s surrounding which help it better in the crowded environment.