Combined coverage and path planning for field operations

Searching an optimal path to cover the whole field while fulfil certain agricultural field operation has been found to be extremely difficult problem to solve mathematically or algorithmically. Although the area coverage problem has been studied extensively in the robotics literature, most of the developed approaches cannot be used for the case of agricultural field operations due to the special characteristics inherent in these operations. In this paper, a first attempt to connect two complementary approaches on the field coverage planning for agricultural machines that have been independently developed is presented. The first approach, using prediction and exhaustive search methods, results in the optimum decomposition of a complex-geometry field into sub-fields and the optimum driving direction in each field. The second one, using a heuristic combinatorial optimization algorithm, results in the optimal sequence that the machine visits the sub-fields and the optimal traversal sequences of parallel field tracks for each sub-field. As an implementation of the total method, an example of optimal planning for a given field is given. Based on this preliminary work, it seems that the resulted combined approach provides a complete method for field area coverage planning that is directly applicable to autonomous agricultural machines as well as to a next generation of navigation-aid and auto-steering systems.