Connected field robots ease the burden for farmers

Weeding robots Oz, Dino and Ted are designed to make life easier for farmers while accelerating the green transition in the countryside. Naïo Technologies, a pioneering French AgTech start-up, has developed these autonomous connected machines.

“Thanks to deep learning, the robot is able to recognise and “cut out” a lettuce in order to adapt the route of its weeding tool.”

For millennia, agricultural revolutions have had decisive impacts on the life and well-being of human societies. Gradual improvements in knowledge, tools and techniques, and occasional technological breakthroughs, have made it possible to feed larger populations: the development of the plough around the 1st millennium BC; the rise of windmills around 1000 AD; the techniques of the British agricultural revolution beginning in the 17th century; and, in the 20th century, the mechanisation of labour with the tractor, coupled with a leap forward in agronomic science (chemical fertilisers, genetics). Today’s advances in computing, electronics and networks suggest that a new agricultural revolution is on the horizon in the medium term.

A revolution that began in the Landes

Following in the footsteps of the “city robot”, found almost everywhere in factories, comes its cousin, the field robot. So goes the legend, or the success story, recounted by Naïo Technologies. The French company is a pioneer in the agricultural robots sector. Over 200 of its machines are now in service around the world: in France, Denmark, Spain, Japan and the United States. The venture began at the Fête de l’Asperge de Pontonx-sur-l’Adour, (Pontonx-sur-l’Adour asparagus festival), which takes place every year in the Landes. In 2010, two food-loving robotics engineers had a chance meeting there with a farmer, who told them about the problems he faced in his work, and something clicked. After months of testing and multiple prototypes, in 2013 they developed the first version of the Oz robot, a small autonomous machine for weeding vegetable plots. Next came the Dino “straddle” robot, specially designed for open-field vegetable crops in both raised vegetable beds and rows (lettuces, carrots, onions), and more recently Ted, for work in vineyards.

A range of technologies

The company designs, programmes, prototypes and assembles its own robots. While the design of the machines varies to suit the different types of crop, the principle remains the same: a fully autonomous electric battery-powered vehicle including a hoeing tool, robust enough to work outdoors in all weather conditions, and equipped with a camera, LiDAR sensor, safety sensor and “bumper” to stop it if there is an obstacle in its way. Before customers’ robots are put into service, their plots are mapped directly with the robot, by surveying and importing existing maps directly onto the seeder, or by drone. The machine then works on its own, without supervision, thanks to a GPS NRTK (the “network” variant of RTKtechnology) guidance system, with centimetre-level accuracy.

Joan Andreu, Director of R&D at Naïo, explains this blend of technology: “Our robots are not yet capable of decision-making as such, but they are active in the field. Thanks to deep learning algorithms, the robot is able to “cut out” the lettuce or vine stock by itself, and adapt the use of its tool to the bumps and holes in the field that are not visible on the map.”

Weeding made easy

This autonomy is the field robot’s greatest advantage. “We are not trying to reinvent the tractor or replace farmers”, says Joan Andreu. “Our robots respond to specific needs that were not being met before. Their primary goal is to save farmers from weeding, which is one of the most time-consuming and physically demanding tasks on a farm.” According to the company, improved working conditions and time savings quickly make the initial investment worth it, and farmers who have opted for this type of machine love it.

This new form of robotics also offers a solution in regions with an insufficient skilled workforce. According to Joan Andreu, “people sometimes forget that agricultural work is highly technical and requires specialist skills. But there are places where, in addition to the complexity of managing seasonal workloads, farmers may find it difficult to recruit, for example, an operator who is familiar with driving a tractor in vineyards.”

Fewer chemical inputs

One final advantage of the agricultural robot relates to environmental issues. With consumers’ growing appetite for “organic” products, and tougher regulations on chemical inputs, “hand weeding” is becoming a critical issue for many farms. “The environmental impact of our robots is not zero, of course, but we ‘go easy’”, explains Joan Andreu. “Even with its battery, a Ted or Dino straddle weighs much less than a four-tonne diesel tractor carrying two tonnes of weedkiller. This reduces the problem of soil compaction and, therefore, water consumption for irrigation. And that means it can cover crop rows more frequently, with greater responsiveness to weather windows.”

This allows for weeding with zero CO2 emissions in use, less oil consumption and easier maintenance. A well-weeded, healthy plot will require less treatment: “Early studies show that our robots are twice as efficient as traditional mechanical processes”, says Joan Andreu. “We certainly do not have an answer to all environmental issues, but if we can help a farmer to reduce the use of chemical inputs by just 10%, it will have a huge impact.”

Data: a key challenge

What about connectivity issues? Naïo Technologies currently bases its products primarily on the 4G standard and is considering the potential of 5G as part of the Orange 5G Lab. “Our robots need a good connection to operate, of course”, says Joan Andreu. “To compensate for the GPS distortions inherent in space triangulation, but also to provide data, on the condition of their systems, battery levels etc. and, in the future, to develop towards fleet management. So we are open to all innovations in this area, especially as our robots, by definition, operate first and foremost in rural, sometimes very isolated, areas that are often poorly covered by traditional networks primarily designed for the urban world.”

An AgTech market worth $18 bn in 2025

The company is expanding while continuing to improve its robots and their maintenance. It is working on new applications in close collaboration with chambers of agriculture and large agricultural groups. It contributes to biological and agronomic research with researchers interested in the opportunity to have permanent sensors in the field to collect data on plants, soils and climates.

Beyond Naïo, a global industry is emerging and forming at the meeting point between agriculture and high technology. The start-up is behind the International Forum of Agricultural Robotics (Forum International de la Robotique Agricole – FIRA), which has become a key stand-alone event, offering robot demos, a scientific symposium, and opportunities for everyone involved in this promising sector to meet. According to the association that organises the Forum, the market is now worth $8 billion, and could exceed $18 billion in 2025. While dairy processors make up the bulk of the current fleet, other applications are on the increase, with 492 robots already in operation worldwide for crop production. There is an abundance of innovation in many areas: autonomous tractors, automated tools, crop dusting, interfaces, collaboration between humans and machines etc.

Despite Covid-19, FIRA 2020 will take place from 8–10 December thanks to an online platform, allowing participants to take part live or via podcast.

Read also on Hello Future

Soft Robotics Lab – ETH Zürich (lab head: Prof. Robert Katzschmann (not in the picture). From left to right: Jose Greminger (Master student), Pablo Paniagua (Master student), Jakob Schreiner (visiting PhD student), Aiste Balciunaite (PhD student), Miriam Filippi (Established researcher), and Asia Badolato (PhD student).

When will we see living robots? The challenges facing biohybrid robotics

Discover
person wearing bioelectronic fibre arrays for dual-ECG signal acquisition / credit: Wenyu Wang and Yuan Shui

Bioelectronics: disease monitoring sensors that can be printed directly onto human skin

Discover

Artificial pollination: robotic solutions that aim to supplement the work of bees

Discover

Autonomous cars: the five levels of autonomy

Discover

Autonomous vehicles may soon benefit from 100 times faster neuromorphic cameras

Discover
GettyImages - Warwick AgTech

Robots developed by Warwick AgriTech could cut herbicide use by 95%

Discover

IoT: start-up ONiO has built a batteryless microcontroller

Discover

Smartphone Collision Alerts for Pedestrians in Paris

Discover