Remote supervision of autonomous agricultural machines

点击翻译按钮获取中文摘要

Automation is being introduced into machines in many industry sectors. Autonomous machines are used in manufacturing. There is growing interest in driverless cars on our streets and highways. Airplanes are flown using an autopilot function. We send drones to investigate dangerous situations. Furthermore, the "Internet of Things" concept has encouraged us to consider how devices can be interconnected, ultimately enabling machines to communicate with each other to function as a team. The age-old profession of farming is also beginning to experience the influence of automation. The concept of the driverless tractor has appeared in the scientific literature over the past couple of decades. A quick search of the internet will produce videos showing driverless tractors moving through fields. Advances in technology are taking us closer and closer to the reality of a driverless tractor, but there are still research issues to be overcome. As has been observed in previous applications with automation, it is reasonable to assume that autonomous agricultural machines will not completely displace the need for the human worker. If designed appropriately, it should be possible for one farm worker to supervise multiple autonomous machines remotely from a central location. The ultimate productivity of such a human-autonomy system will depend to a large extent on the ability of the farm worker to efficiently and effectively obtain information from the autonomous machines through an interface. This implies that great importance must be placed on the design of the automation interface so that human interaction with the autonomous machine can be optimized. There is a current void in the scientific literature with respect to the design of interfaces to support shared understanding among teams comprised of both humans and autonomous machines. The long-term goal of this research program is to generate knowledge that improves our understanding of the human-machine interactions associated with the use of agricultural machines. Over the immediate five-year period, the focus is to understand the issues associated with remote supervision of autonomous agricultural machines to enable the design of an interface that facilitates the sharing of information between the autonomous machine and the farm supervisor. The proposed research will have an immediate impact on the agricultural industry as domestic and international companies are currently developing autonomous tractors and will require the knowledge generated by this research to successfully commercialize their technology. Knowledge gained will be relevant well beyond the agricultural sector as remote supervision of autonomous machines is expected to be used in numerous industries. It is proposed that 2 PhD, 3 MSc and 5 UG students will be trained through this research. They will become experts in evaluating human-machine interactions for autonomous machines.