Ensuring food safety for a growing global population while addressing the impacts of climate change is one of the key challenges facing modern agriculture. At the same time, labor shortages, increasing efficiency demands, and the need for safer field operations are driving the adoption of new technologies. Autonomous machines and AI-enhanced applications have the potential to help farmers meet these challenges by optimizing resource use and automating complex tasks. Advancing this research was the objective of the recently concluded, EU-funded AGRARSENSE project, which explored innovative approaches to autonomous agricultural machinery and intelligent farming operations.
The AGRARSENSE1 project ran for more than three years – from 2023 to February 2026 - and brought together a consortium of 57 partners (including 4 affiliates) from 15 countries. One of the world’s leaders in forestry automation, Komatsu Forest AB, served as the overall project coordinator.
“AGRARSENSE looked at how sensory and automated capabilities can improve agricultural technology and productivity in Europe for future generations. We were excited to be part of the project and to contribute to two of the seven use cases. TTTECH and TTCONTROL’s locations in Austria and Italy participated jointly, with TTCONTROL taking over the role as the main pilot project partner. This was also the first participation of TTCONTROL Srl in Italy in an EU-funded project,” says Martijn Rooker, Senior Innovation Projects & Funding Manager at TTTECH.
Pilot use cases: TTCONTROL provides high-performance computing capability
TTCONTROL contributed to the agri robotics and forestry pilot projects, providing performance computing platforms for the applications developed. These platforms are essential in more autonomous operations, serving as the “brain” of the vehicle architecture coordinating the ‘central nervous system’ consisting of sensors and cameras that collect data about the machine’s surroundings and the applications that will merge the different data sources and use that data to optimize the behaviors of the systems being tested.
High-performance computing capabilities are a key requirement to be able to process the large amounts of data generated and to run AI-powered algorithms for fast and efficient decision-making and actionability.
In the pilot use cases, TTCONTROL’s 25 years of off-highway market expertise across a range of sectors informed their contribution, according to Martijn Rooker: “As a project partner, TTCONTROL brought their implementation, software, and application development experience to the table. This supported the successful integration of their robust performance computing platforms that enable a range of AI-enhanced services into the two pilots. It was a great opportunity for us to test these solutions in the field in very specific settings.”
Autonomously operating agri robotics for vineyards
The use case led by Agreenculture SAS focused on an autonomous agri robotic solution for perennial crops such as vineyards, consisting of a robotic platform and a trailer with a robotic arm implement for pruning that can cut grape clusters. It uses cameras to monitor the quality of grapes and provide visualization of the pruning process to the vineyard owner. To ensure safe operation, it uses motion sensors and thermal cameras to detect people and animals and avoid collisions.
TTCONTROL’s performance computing platform was mounted to the robotic platform from Agreenculture, which served as the pulling machine. It received and processed the camera and sensor information directly on the machine, without needing external computing resources. This edge computing capability is necessary to allow the robot to autonomously decide whether a vine needs to be pruned and engage the cutting process, as well as to determine which information is relevant to be transmitted back to the vineyard owner via a wireless communication system.
Autonomous driving for forestry application
Project coordinator Komatsu Forest AB was the partner in this use case which aimed at improving the tree harvesting process. The process requires two machines with one operator each – a harvester that cuts the trees and the forwarder that grips them for transport. The goal was to find a solution that would allow the use of smaller, autonomous shuttles running more often and with smaller loads instead of the unwieldy and heavy forwarders that put a strain on the forest floor.
TTCONTROL’s performance computing platform was used for processing sensor and camera data and running the autonomous driving solution of another pilot partner, which was used for object detection and calculation of a safe driving path. The operator sat in the cockpit but no longer had to drive themselves.
The base setups for autonomous applications are often similar: They begin with sensors and cameras that monitor the surroundings and collect data. This is then sent to a central computing platform for processing and AI-driven decision making. From there, the actions to be taken are deployed to the system/equipment executing the prescribed actions. However, each use case is different, depending on the machines used, application complexity, as well as industry-specific standards and requirements.
What makes research projects like AGRARSENSE so valuable for TTCONTROL is the testing grounds and the close cooperation with manufacturers, research institutions, and industrial companies, as Martijn Rooker puts it: “Research projects are fascinating and worthwhile because we get together with experts from different disciplines working towards a common goal. The ecosystem approach drives forward a lot of ideas and companies can then test them and build upon them for enhancing their solutions portfolio. In AGRARSENSE, we were able to further develop our platform solution and to see how much data is needed, what kind of data can be collected, and how our performance computing platforms reacted in real-world environments.”
1 AGRARSENSE receives funding from the European Horizon Europe research and innovation programme, KDT Joint Undertaking (grant agreement no. 101095835) and from the Italian Ministry of Economic Development (MISE funding).
Find out more
- Project website: AGRARSENSE
- Advancing smart farming: TTCONTROL collaborates in aerOS project pilot
- HORSCH uses TTCONTROL's computing platform to further develop AI functionality in its spot spraying solution
- How OEMs can successfully integrate AI into off-highway applications
- Performance computing platforms that can be leveraged for agricultural, forestry, municipal, or special vehicle use cases.
- THEIA-XR: Advancing operator assistance and safety in off-highway machinery with extended reality
