Eureka! Institutional Repository

One of the most evolving state-of-the-art technological solutions is the development of Intelligent Autonomous Vehicles (IAVs) that navigate on industrial facilities taking real time decisions with minimal or zero human assistance. IAVs rely on microelectronics and sensor technologies for monitoring, communicating, coordinating with the facility’s equipment or with other vehicles.Our custom built IAV is a prototype for industrial facilities that can monitor industrial processes in a facility and enhance the decision making process in the shop floor. For our prototype we used a handmade chassis with two separate geared motor-driven wheels along with encoders to control their revolutions and a single caster wheel. The main data processing unit is a Raspberry Pi board, communicating through a standard Wifi shield and a PiFace digital 2 extension board with relays. Additionally, we included an L298N dual H-bridge to drive the motors, a camera for object identification and a LiDAR laser sensor for obstacle avoidance, area mapping and identification of regions of interest. Moreover, we added sufficient cooling with a 12mm fan, an adjustable switching regulator to power the electronics, and finally, a Lithium Polymer (Li-Po) battery to power everything. The PiFace is a H.A.T. (Hardware Attached on Top) board, which is used with the Raspberry Pi and consequently cuts off access to the original pins. In our case, we created a custom board, and attached the PiFace on top of it, releasing the majority of the pins for further use, as will be shown afterwards. Furthermore, we developed a custom operating system (OS) image for the Raspberry Pi based on Ubuntu Linux and the Robot Operating System (ROS). ROS is an open-source, meta-operating system that provides hardware abstraction, low-level device control, data fusion and manipulation as well as libraries for building and executing operations across multiple entities. Our vehicle can be remotely controlled with the camera node and a custom ROS interface while on the same time identify the surroundings of the facility layout using the LiDAR laser sensor. The vehicle could autonomously navigate in an industrial facility and monitor the real time processes in order to provide decisions maker with valuable feedback from the shop floor.