Tractors face the risk of lateral rollover while cornering on slopes. Coping the situation is important for safety. In this study, dynamic equations were formulated for four degrees of freedom (vertical, pitching, rolling, and horizontal), while considering the centrifugal force. The centrifugal force affecting the tractor body was verified through a modeling experiment. Furthermore, tractor cornering behavior on a smooth slope was simulated. Consequently, the theoretical values and the measured values of the centrifugal force matched less than 4.0 m/s. Additionally, the cornering tractor overturned on a gentle slope (3°-8°) at approximately 4.0 m/s² centripetal acceleration. Aforementioned results indicate that to prevent overturning, even on a gentle slope, it is necessary to minimize the traveling speed.

In order to navigate a robotic combine harvester, we have developed a rice field image segmentation methods based on deep learning architectures. Based on several classic convolutional neural network models, eight different segmentation models were constructed. After training, the speed and accuracy of all models processing the same dataset were analyzed and compared. The results indicated that the entire group of models were able to effectively perform the segmentation of the harvested areas, unharvested areas, ridge areas, and background areas in the images when the unmanned rice combine harvester was in operation. The pixel accuracy, class mean accuracy and mean IU (intersection over union) of the best model were 95.17 %, 86.08 %, and 80.07 %, respectively.

The aim of this research was to simplify the suspension system of running device on a one-wheel rice transplanter. The free-pivoting suspension of the wheel was mounted on a modified rice transplanter with a fixation device installed for switching between the free-pivoting and the fixed suspensions. In varying soil pans of the field, the pitch angle of the transplanter with the free-pivoting suspension was maintained stably at an average of 1.8° on the soil surface, compared with 4.7° that tended to change with the sinkage of the wheel with the fixed suspension. With the addition of 98 N of weight on the wheel, slippage with the free-pivoting suspension was reduced by a half on uniform surfaces of the soil pan.

We assessed the environmental burden of strawberries using life cycle assessment (LCA) with the damage to these products due to vibrations during transport. The vibration test for the transport process was performed for several transportation distances. This test allows the determination of the ratios of damage area as an indicator for injury to strawberries during transport. The environmental burden of the packaging scenario was compared with that of the non-packaging scenario. The cultivation process was a large contributor (81.2～99.1 %) to the environmental burden (and thus, a major portion of the life cycle). The environmental burden of the packaging scenario was lower than that of the non-packaging scenario, indicating that package can reduce food loss during transport.

The tractor engine power in rotary tillage for calculating fuel consumption was predicted by the simplified estimation method in Hokkaido soil. The PTO shaft output power was predicted from the PTO shaft output energy per tillage volume (kWh/m³). The specific drawbar power that means the ratio of the drawbar power to the engine power was classfied the typical traction efficiency in the different soil series and soil surface conditions. The engine power values were summed up as the PTO shaft output power that was calculated from the physical soil characteristics and the drawbar power that was calculated from the specific drawbar power. A linear regression equation of this engine power versus the engine power calculated from the fuel consumption in the dynamometer test had a high prediction accuracy with the slope of the regression equation around 1.0 except that the fuel consumption engine power was 3 kW higher than that in the dynamometer test. This prediction method has no pratical ploblem in the lowland soil of Sorati region and the andosol soil of Tokachi region.