Technology Assessment of AI-Based Computer Vision for Harvesters in Industrial Mixed Crop Production

Conventional agriculture faces escalating environmental and economic challenges. Intensive monoculture farming and the extensive use of chemical additives, such as fertilizers and pesticides, contribute to environmental pollution, biodiversity loss, and soil and water degradation, threatening both natural resources and the long-term resilience of agricultural systems. Mixed cropping, the cultivation of multiple crop species together, presents a promising alternative by optimizing resource use, enhancing soil and water quality, and reducing dependency on chemical inputs. However, its industrial application remains limited due to complexities and the lack of technical solutions, particularly in harvesting. Computer vision technology holds great potential to address these technical limitations. This paper supports agricultural machinery manufacturers by presenting a technology assessment of computer vision applications in harvesting machines for industrial mixed crop cultivation. Functional capabilities of computer vision were categorized and matched against the requirements of various functions within industrial mixed-crop harvesting, creating impact matrices that evaluate the relevance of computer vision per function. The results demonstrate the substantial potential of computer vision to facilitate, or even enable, the harvesting of mixed cropping systems at an industrial scale. The findings offer a technological foundation for developing mixed crop harvesters, informing evaluation and selection processes for final design solutions.