AVL Develops ‘Generic’ Automation for Off-Highway Vehicles

March 24, 2026

Bernhard Knauder and Johannes Roth

Conventional series tractor equipped with GOAS in real-world field testing. *(AVL List GmbH)*

Off-highway equipment manufacturers face increasing pressure to deliver safer, more reliable and more efficient machine operations while navigating workforce shortages and rising productivity demands. Autonomous and automated vehicle technologies offer a pathway toward meeting these challenges, especially in agricultural and construction applications where repetitive tasks and large operating areas can be systematically automated and optimized using precision technologies.

The Generic Off-Road Automation System (GOAS) was developed by AVL to provide a modular, vehicle‑agnostic approach to automation, enabling manufacturers to integrate autonomous capabilities into both new and existing platforms without requiring major architectural redesigns. GOAS differs from previous automation concepts by using a skill‑based architecture that decomposes autonomous operation into modular functions.

Earlier systems were typically developed for a specific machine type designed for specific working modes, making cross‑platform use costly or infeasible. In contrast, GOAS defines autonomy through reusable capabilities – such as navigating between points, following a predefined path, or executing a tool‑specific task – allowing the same software modules to operate across tractors, construction machines or mining vehicles.

The seamless configuration of the automation kit function range, supporting various applications, is enabled through clearly structured and open interfaces on the outer as well as on the inner system components. A specific focus is given on the capability to retrofit machines, which are already operating in the field, providing an automation upgrade with minimum need for system adaptations. The result is a flexible, scalable architecture that shortens integration timelines and lowers engineering overhead.

Shift-left development using high-fidelity simulation

Overview of the autonomy kit system boundaries, scope of included functions, and interactions with the vehicle. (AVL List GmbH)

Simulation plays a central role in GOAS development. Using a shift‑left methodology, engineers evaluate control algorithms, perception pipelines and planning behaviors within a virtual environment before physical testing begins. High‑fidelity vehicle dynamics models replicate engine response, steering characteristics and tire‑soil interactions, while 3D simulated fields allow perception systems to process realistic crop structures, lighting changes and obstacle scenarios. This enables consistent testing across sensor configurations and reduces reliance on weather‑dependent field trials. Furthermore, the speed of development is boosted by the possibility to validate the control system in early project phases under realistic test conditions, as well as the possibility to support field testing activities by replication of experienced scenarios.

Phased integration methodology

Three phases of automation system integration from CBW to AD to ATE. *(AVL List GmbH)*

GOAS uses a three‑phase integration framework to structure the path from basic actuation to full mission automation. The first phase, Control‑by‑Wire (CBW), establishes electronic control of vehicle motion (steering, propulsion, braking) and relevant functions for working task execution (e.g. PTO, hitch, hydraulic valves).

Once CBW is validated through simulation and real‑world testing, Phase 1 introduces Automated Driving (AD) features such as path following, obstacle handling and navigation within a defined operational design domain. Phase 2 extends the system to Automated Task Execution (ATE), enabling the vehicle to perform complex workflows – such as field coverage or repetitive transport cycles – without operator intervention.

Real-world case studies

Two vehicle platforms were used to validate the system. The first platform represents an electric tractor prototype equipped with lidar, radar, cameras and combined GNSS/IMU positioning. Simulation allowed engineers to validate perception and control logic under diverse conditions before transitioning to physical testing.

Prototype electric tractor equipped with GOAS, featuring lidar, radar, cameras and GPS/IMU sensors for autonomous testing and task execution. *(AVL List GmbH)*

The second platform is a conventional series tractor retrofitted with the same sensor stack and control interfaces. Intentionally, the series tractor was chosen to be a model that has been available on the market for more than 10 years, proving the retrofitting capability of the automation system concept.

Despite structural differences, both vehicles achieved stable autonomous driving behavior and reliable execution of task‑level functions, demonstrating GOAS’s cross‑platform applicability.

Conclusion and outlook

GOAS demonstrates that scalable autonomy for off‑highway vehicles is achievable through modular design, simulation‑driven development, and a phased integration approach. By separating autonomy into reusable skills and validating early through virtual workflows, manufacturers can reduce development cycles and integrate automation into both new and existing machinery.

Future development work will focus on expanded field validation, integration into multi‑machine fleets and continued enhancements in functional safety to support large‑scale commercial deployment.

Bernhard Knauder, Skill Team Leader E/E, SW, Controls and Calibration, and Johannes Roth, Director of Product and Business Development Off-Road, AVL List GmbH, Steyr, Austria, wrote this article for SAE Media.