Published On 03 January 2024
We are pleased to present the second episode of Meet Our Tech, highlighting BEEYONDERS’ breakthroughs in autonomous and teleoperated ground and aerial vehicles.
Construction sites, known for their dynamic and hazardous nature, demand robust perception and control systems for autonomous ground vehicles. BEEYONDERS addresses this by pioneering an open-source software platform for autonomous machinery. This platform guides heavy vehicles and drones with equal efficiency.
Equipped with multi-sensor perception, these vehicles generate accurate 3D maps and share them in real time. The software updates these maps on the fly, responding to environmental changes not reflected in the existing maps. To achieve this, a high-level decision-making module is under development. It will coordinate and schedule tasks among machines, learning from the process using a reinforcement learning approach focused on efficiency and productivity.
In environments where a GPS is unavailable, aerial drones will operate autonomously using 3D Lidar technology. This technology, combined with inertial sensors, enables autonomous navigation in GNSS-denied environments and enhances obstacle identification and avoidance capabilities.
When asked “What is the greatest challenge in developing these technologies?” Maria Teresa Lazaro and Carlos Bernad, experts from Itainnova, gave this answer:
“Developing autonomous machines for tunnel and construction sites involves addressing critical challenges arising from their unpredictable and dynamic nature. These environments demand adaptive systems capable of handling diverse obstacles and rapidly changing conditions.
Additionally, factors like low visibility and dust further complicate navigation, prompting us to implement advanced perception sensors including radar and specialised cameras such as thermal, or infrared ones. These non-standard perception sensors play a key role in mitigating the impact of low visibility and dusty conditions. By integrating these technologies, our autonomous vehicles can ensure accurate perception, enabling safe and efficient navigation within tunnels and construction sites.”
The machines should operate together without human intervention while performing close to a human-equivalent: for now, the two great ambitions of the project are to optimise task sequencing for more efficient use of resources for ground vehicles, and make possible autonomous flight for drones in buildings or low visibility conditions.
Prototypes of these robotic platforms with autonomous functionalities will undergo testing in hazardous environments to validate their field functionality. BEEYONDERS’ case studies will include:
• Santa Barbara Foundation Test Tunnel (Spain): Drones will scout rocky and debris-obscured terrain in an underground passage after explosions.
• AISCAT 5,000 km Highway Network (Italy): Autonomous vehicles will optimise pathways, automate maintenance and inspection of traffic lights and signals, and monitor vegetation growth beside roads.
• Building Construction in Helsinki’s Metropolitan Area (Finland): Drones will gather end-of-day progress information.
Success in these endeavours promises a safer and more resource-efficient work environment. Stay tuned for more updates on all BEEYONDERS technologies!