University researchers have developed a low cost device that can be adapted to different types of vehicles, allowing the movement and transport of loads in complex spaces without the need for drivers. This system is ideal for transporting materials in warehouses where it is necessary to make a continuous passage of goods between different points.
The technology, designed by researchers from the UA Research Group in Signals, Systems and Telecommunication, allows the vehicles to be programmed to travel independently, without human intervention, by mapping and pinpointing the area through various sensors such as laser and machine vision systems. This allows for the reconstruction of the area and the identification of possible moving objects. It also reckons the optimal path to reach a certain point, avoiding potential obstacles through reinforcement learning techniques.
The automatically guided vehicle (AGV), heretofore used in warehouse management, are based on the use of previously established paths, that is, guided by painted lines or lasers. These systems have limited flexibility and no ability to solve unforeseen situations, which can lead to unexpected objects that hinder the movement of goods and obstruct the passing. Furthermore, the establishment of new routes involves programming prior work with high temporal and economic cost. Ph.D. Tomás Martínez Marín, main researcher of the groups explained that the intention was to convert any manually driven vehicle into a high-performed mobile robot to suit the working environment where it is going to operate, rather than having to adapt the environment to the vehicle, with the high cost of doing so.
This research work has led to a technology and a number of algorithms patented by the University that improve processes mapping, association and location of robots over SLAM conventional techniques (Simultaneous Localization And Mapping).
This technology is applicable to any business or industrial environment where autonomous vehicles can be used, for example, in the management of warehouses with large volume of inputs and outputs, production centres in need of displacing objects between different points, movement of goods in settings with extreme human conditions (cold storage, waste management , etc….)
The device is available to be adapted to conventional vehicles. As a practical demonstration of this technology, a golf cart has been automated and it is currently circulating independently on the UA campus. Also, there is a virtual reality simulator to visualise the application of this technology in specific tasks and facilities of each company. Therefore, the research group has the capacity to develop a comprehensive system of control of a fleet of vehicles based on the needs of the company.
Cite This Page: