Advanced levels of autonomy and intelligence in agricultural and forestry utility vehicles are imperative for the solutions to global as well local challenges with particular focus on agricultural production system in this doctoral thesis. 

Historically, it is seen that the application of scientific and technological developments through increased mechanization and precision farming has provided several opportunities for agricultural production in general and within forage handling operations in particular along with similar field operations in forestry. Some promising engineering developments in the 20th century with regards to forage handling include forage harvesters, balers, and automated wrapping equipment of balers using stretch films of 25µm thick to lower risks of dust, molds, spores, mycotoxins respiratory allergenic disorders in livestock and humans. Baler machines have made it possible to trade silage (harvest and storage of moist grass using fermentation) in portable packages between farms which typically weigh 600-800 kg freshly cut per bale and are more popular on smaller farms with limited labor and financial resources to construct silos. 

Bales made up of hay or silage formed by balers are usually too heavy to be picked up by humans alone. Thus, they are picked up from fields using conventional utility vehicles such as tractors or loaders operated by a human. These kinds of operations are labor intensive and associated with health and accidents risks. There is also a potential to further improve efficiency and environmental impact since most decisions are made by humans and thus limited to human labor capabilities in terms of load handling, sensing, multitasking, planning, consequence analysis etc. One significant contribution could be relieving the human operator from tedious driving tasks and incorporating optimized automated planning and autonomous navigation functionality for field operations such as haybale collection.

The prime focus of this doctoral thesis is to advance the existing autonomy level and intelligence in agricultural vehicles for risky and labor-intensive field operations while demonstrating a proof of concept (PoC) in pertinent forestry operations. This began with investigating current challenges and opportunities with agricultural vehicular automation and potential improvement for one of the field operations. Haybales collection operation is found to be one of the riskiest operations and taken as one case with great potential for improvement with automation. As result, proposed innovative solution based on the new study of path planning approaches for bales collection operation in fields of varying complexity shows that optimized solution with new concept crane-vehicle can produce 20-30% reduction in working distance for this task. To complement further, the full-scale autonomous navigation control system based on heading error correction and different reference points for articulated vehicles are developed, evaluated and tested.,  Pure pursuit algorithm with front-axle midpoint (PPF) produces less cross track and pure pursuit algorithm with virtual midpoint (PPV) produces less heading errors for innovatively designed articulated or center-steered vehicle(AORO) hence making it suitable choice for operations where heading errors are of concern such as haybales collection or similar tasks. Together, innovative optimized planning solution & improved autonomous off-road navigation control developed in this doctoral thesis showed 20-30% distance reduction and up to 50% reduction in real-time tracking errors respectively in complex unstructured real environment; when combined translating into the high plausibility for overall greater safety, high productivity and efficiency for agricultural production system in particular with wider implication in pertinent field operations such as in forestry.