Automation in agricultural machinery is a crucial driver of productivity and sustainability. Some automation features like automated steering and real-time data analytics are already state-of-the-art. On the other hand, a human driver performs the optimization of the working speed manually, and the automation of this is an ongoing challenge. Process quality and process efficiency are the two main targets in this optimization. Agricultural soil tillage requires achieving both. Therefore, the correlation between process quality optimization and process efficiency is fundamental, and vice versa. The approach presented in this paper shows how the two optimization targets of efficiency and process quality can be optimized and aligned together. Optical sensors determine various parameters to describe and model the process quality. The measured machine state determines the characteristics of the interaction forces between the machine and the environment. A machine learning algorithm describes the relationships in the drivetrain. The two process targets are each predicted for different working speeds and are combined in the form of a boundary target and an optimization target to identify one optimized target speed value.

The requirements for modern drivetrains are increasing across all industries. Even mobile working machines such as agricultural and construction machinery are subject to increasingly higher demands in terms of efficiency and CO2 emissions. To verify these requirements and drive further development, it is necessary for testing processes to comprehensively evaluate the machine and its operational processes. For this purpose, the MOBiL testing approach was developed at the Institute of Mobile Machines. This approach incorporates parallel drivetrains, information flow and the environment of the driving and working task. To implement this approach in a complete vehicle testbench, a framework was developed that enables fully individual driving and working tasks of a mobile working machine to be replicated on a test bench. The basis for this framework is the Robot Operating System (ROS), which runs various nodes. Individual nodes control the different testing subsystem, such as the 4-WD-Acoustic-Chassis Dynamometer, the driving robot, or a PTO test stand. Starting from a simulation node, the subsystems are controlled and reflect their respective measured variables back into the system. These variables are then returned to the simulation and collected along with data from the simulation. All nodes are interchangeable, allowing suitable specifications, simulations, and hardware to be used depending on the scenario: For instance, a tractor with a PTO-driven implement like a baler or a rotary harrow driving on a field. This allows tests on the powertrain and its controls to be reproducibly researched on a test bench. Therefore, efficient, and customized testing processes for a wide range of applications and vehicles are enabled.

Accurate localization is one of the key requirements for the automation of mobile machines. While GNSS-based systems are widely used due to their high accuracy and accessibility, redundant systems have to be developed to decrease the dependency on GNSS signals for autonomous machines. Although altitude measurements have been used for many decades by human explorers, they are not yet exploited for localization purposes on mobile machines. Based on single barometric measurements, no localization is possible as the vehicle could be at various positions with the same absolute altitude. In this paper, we propose an algorithm based on sequential importance sampling to fuse altitude measurements with surface profiles, which allows real-time tracking and localization of mobile machines. When moving through hilly terrain, a machine constantly changes its altitude, and the altitude profile can be used to track the vehicle's position. The proposed algorithm offers a supplementary localization and verification method for mountainous and potentially GNSS-obstructed areas.