Challenges and Solutions

Agricultural robots have emerged as innovative and eco-friendly
means for easier farming. In particular, autonomous weeding allows farmers to efficiently maintain their crops without using herbicides or chemical weed control products. Electrically-driven robots require less energy than a tractor. They are also lighter than a tractor and thus avoid compacting the soil, even when used frequently.

Naïo Technologies is a pioneering SME in this domain. Their first product was Oz, a weeding robot for small crops of vegetables. The company is now targeting the market of large-scale vegetable crops and vineyards, and has started the development of bigger and faster robots. However, the deployment of such robots introduces new potential risks to the users and their farm. Since the safety regulations and standards are not yet established for autonomous robots in agriculture, SMEs like Naïo
need to adopt a proactive approach in the acquisition of dependability-related technologies. In order to implement a successful solution Naïo needed skills concerning safety analysis and validation technologies.
More precisely, Naïo increased its ability in performing structured and systematic analysis of operational risks, formal assessment of architectures and simulation-based testing.

DIH Support

The DIH provided Naïo with access to leading edge safety methods such as HAZOP-UML for the analysis of the operational risks induced by the robots and Altarica for the assessment of candidate robot architectures. It also
helped Naïo to establish simulation-based testing as an essential part of their validation process, based on recent research results on testing robots in virtual worlds. The DIH facilitated the collaboration of its members
with Naïo by not only providing academic tutorials but also by focusing on practical problems and bridging the gap from robot prototypes to mature products.

Impact/What’s next

An immediate impact of the experiment was a one-third reduction in validation costs, for the functionalities that can be tested with the developed simulator. Naïo now use this simulator for two robots out of the four they have in development. In the longer-term, the transfer of safety-related technologies has put Naïo into a better position to get into the market of large-scale crop maintenance. Meeting legal requirements, in particular as regards safety, is a prerequisite for selling their robots at the European and international levels. The experiment was part of Naïo’s continuous effort to reach the business objective of doubling their turnover
each year. They expect to go from 1.2M€ this year and 40 robots sold, to more than 10M€ of turnover and 300 robots in 2020. This ambitious objective is in line with the huge potential for development in agricultural
robotics. Tractica, a US research organization, forecasts that the agricultural robot market will increase exponentially from $3 billion in 2015 to $16 billion in 2020 and then $73 billion in 2024.