Forgotten by its masters, a little clean-up robot is working on the surface of the devastated planet Earth. Powered by solar energy, Wall-E piles up the garbage left behind by our civilisation. In Pixar’s animated film of this name, which came to our screens in 2008, the robot was struggling to stay alive in a post-apocalyptic universe where the Blue Planet, turned into a garbage dump by excessive consumption, had been abandoned by the human race. But clearly, prevention is better than cure. In contrast to this fictional universe, some pioneers today are taking a far more optimistic approach to deploying robots and drones, programming them for environmental protection duties with the intention of preventing the catastrophic Wall-E scenario from becoming reality.  

robots for agriculture

naio technologies

The field where the most concrete action is currently taking place is agriculture. A broad range of startups began two or three years ago to develop autonomous vehicles capable of either limiting or completely eradicating the use of herbicides. These solutions have come just at the right time as the farming world is now seriously seeking alternatives to using glyphosate.

In 2011 a pioneering Toulouse-based startup called Naïo Technologies embarked on this journey. The company, which has just completed a second round of funding worth €2 million, currently has around one hundred robots in the field. With turnover quadrupling and the workforce doubling in the space of two years, Naïo Technologies has enjoyed very strong growth. As with many of its rivals, the ‘raison d'être’ of the company is its environmental aspirations. “We’re aiming to respond  to the challenges in the agricultural sector in terms of both manpower and productivity, but we intend to do better than what exists today, i.e. we want to support and assist farmers in their quest to reduce their impact on the environment,” declares co-founder Aymeric Barthes. That means helping farmers to carry out the more repetitive, time-consuming tasks without having recourse to chemical products.



Autonomous weeding machines now flourishing

Naïo Technologies’ first solution – Oz, a compact vehicle designed to carry out mechanical weeding in small market gardens – was launched on the market in 2013. Last year, Oz was joined by Dino, a machine built for industrial-sized market gardens of ten to a hundred hectares in size. Ted, a weeding robot designed specifically for vineyards, is due to complete the line-up in 2018.

The solution developed by Swiss startup Ecorobotix differs in that it is not a mechanical weeding system. Drawing on Artificial Intelligence, their robot uses a system of what they call “precise detection and discriminating spraying of weeds” so as to considerably reduce the amount of herbicide required. The machine has two tanks, each containing twenty litres, to feed the sprayers placed at the ends of its articulated arms. It is covered with photovoltaic panels, which enable it to keep working continuously for up to twelve hours a day. Other companies, such as Vitirover and Vitibot, that specialise in weeding vineyards, also say they wish to build solutions capable of replacing products that are harmful to the environment. In fact there is no shortage of market opportunities: 70% of all vineyards in France are currently treated with chemical herbicides. The other winegrowers use mechanical methods – i.e. deploying large tractors that risk damaging the vines.

AI  enablING ultra-precise spraying



We’re no longer at the stage of innovation known as ‘technology push’. Instead there is a very real market pull. Today the potential customers are many and varied.
eric marty

Eric Marty

This year, Vitibot is readying the launch of Bakus, a 100% electrical high-clearance mechanical weeder-robot. Vitirover’s robot , driven by solar power and steered by smartphone, is able to navigate with precision between the vine stems in order to remove weeds, within the limits set by its GPS.

Some years ago, these inventions were real technological gambles, but now they are definitely seen to be meeting clear needs. “We’re no longer at the stage of innovation known as ‘technology push’. Instead there is a very real market pull. Today the potential customers are many and varied. And the suppliers of these technologies can clearly see how they can position themselves on the international scene, with the United States and Japan looking to be very promising markets,” explains Eric Marty, Managing Partner for Venture Capital activities at Demeter, a private equity firm specialising in green technologies, underlining: “John Deere’s acquisition of the Californian automated weeding systems firm Blue River Technology [last September - ed] for over $300 million has proved to be a shot in the arm for the sector. Since then we’ve seen a surge in startups working in this field, which has become a target for investment funds.” In France, demand is driven by the need for farmers and winegrowers to reduce their use of crop protection products by 50% by 2025, in line with the Ecophyto 2 Plan published by the Ministry of Agriculture and Food.

ROBOTS and ai moving into selective sorting



Deep learning revolutionising waste sorting

robots by zen robotics


Artificial Intelligence and robotics are also making significant inroads into industrial waste sorting. High-tech recycling centres are starting to install new-generation automated sorting stations equipped with object recognition technologies capable of identifying and sorting garbage efficiently. Unsurprisingly, these robotic waste sorters draw on the ‘deep learning’ approach in order to identify the different materials and direct their articulated arms so as to deposit the waste in the right place. In the long term these machines will replace the traditional sorters which currently use infrared cameras. The cameras have certainly improved the efficiency of the recycling process, but this approach still requires the recycling centre employees to carry out the work as guided by the cameras. Among the new entrants into the waste-sorting business, Finnish company ZenRobotics has popularised the use of robots equipped with high-performance sensors and using machine learning to sort piles of industrial construction material. Some sorting centres have already installed this state-of-the-art equipment. A number of competitors have come up with similar solutions, including US companies Waste Robotics and Bulk Handling Systems (BHS), whose Max-AI robot is equipped with AI technology developed by Barcelona-based startup Sadako Technologies.

cleaning, maintenance and inspection


Robots and Unmanned Aerial Vehicles (UAVs, aka ‘drones’) are also making a contribution to the well-being of the planet in the field of maintenance and inspection of renewable energy production sites. Drones are for example increasingly being deployed to detect surface cracks in the panel arrays at solar energy farms. Analysing the pictures taken by these drones saves a considerable amount of time compared to a traditional site inspection.

Robotics can also be used for cleaning purposes. Solar power generation sites are often located in arid zones swept by winds that carry suspended particulates so the equipment has to be regularly cleaned in order to function at maximum capacity. In the United States, SunPower deploys cleaning robots originally developed by Greenbotics, a Californian company which SunPower acquired in 2013. According to the solar panel manufacturer, which is in turn majority owned by energy giant Total, this automated process enables cleaning to be carried out twenty times faster, and using 75% less water, than manual methods. In the same niche, Israeli company Ecoppia has developed patented technology for water-free panel cleaning.

Meanwhile, other firms are specialising in wind farm supervision. Maintaining the installed equipment normally calls for the scheduling of cumbersome and costly inspection cycles, which are even more onerous for wind farms installed out at sea. Checking the blades means shutting down the turbine and bringing a team of technicians over to the site by boat or helicopter to inspect the blades up at the top of the installation. Sterblue, based in Nantes in western France, and two fledgling British companies – Perceptual Robotics and VulcanUAV – are developing the use of drones as an alternative for inspecting wind farms out at sea. “The use of drones to automate industrial-scale inspection is booming,” confirms Eric Marty. He nevertheless points out that “the sector is becoming highly competitive, which means that each player needs to find a technology differentiator – for example a better problem identification algorithm or a lower cost structure than the competition.”

DRONES for the PLANet

drone des mers

Drones helping to avoid pollution


diya one

A slightly sparser array of initiatives has been launched in other environmental areas such as water purification and air cleaning. The highly publicised Diya One robot from Partnering Robotics, which looks a bit like R2D2 from Star Wars, analyses air quality in hospitals, shopping malls, museums, etc, and then purifies it. It moves around independently, successfully avoiding obstacles. “Robots are particularly useful for measuring air quality in public buildings,” underlines Paul Foucher, Head of Projects at Cleantech Open France, a programme designed to identify and accelerate startups and smaller firms in the ‘CleanTech’ field. “As regards assessing external air pollution, startups such as for example Plume Labs, are focusing on the development of ever-smaller sensors capable of identifying a range of different pollutants,” he explains.

Progress in miniaturisation means that very soon drones will be equipped with these types of sensors in order to track the level of particulates in the air in targeted areas. This is already a reality for the Finnish startup Aeromon, which is currently testing out the implementation of drone-based systems to measure industrial gas emissions. Last April, French integrated oil and gas company Total also set off down this road. Teaming up with researchers based at the University of Reims who are working under the aegis of the French National Centre for Scientific Research (CNRS), the company is working on building a drone capable of assessing with precision the air quality around its refineries. The collaboration between the research team and the energy corporation is centred on developing mini-sensors that are able to model in real time the density of gas in the air, and capable of assessing emissions even at very low thresholds.

metal sharkS cleaning up THE oceans

waste shark

Meanwhile water pollution has now become a pressing issue. To address the problem, a number of companies are starting to develop marine drones that can be deployed to clean up expanses of water. In the Netherlands, startup Ran Marine is using special UAVs to clean up ocean pollution in port areas. Its autonomous machines, known as WasteSharks, are able to gather up to 500 kilos of garbage in their waste baskets. The aim is to capture the floating waste in a port area before it is carried out to sea by the tides and currents. At the same time, the WasteSharks are also programmed to collect data on water quality.

The Jellyfishbot developed by French company IADYS is based on the same principle. The 70-centimetre-long autonomous landing net weighing 16 kilos is able to work for eight hours at a time and can gather up 80 litres of floating debris, including oil, over a surface of some 1000 square metres. The current prototype, which was still in running-in phase at the port of Cassis in the south of France in January, is scheduled for launch on the market in April this year. The bot is activated by remote control, and the company is already working on an autonomous version.

However, these startups, whose technologies are still at the development stage, do not yet have the sort of experience which their counterparts in the agricultural sector – who already have several years of testing under their belts – have built up. It therefore remains to be seen whether this approach to water cleaning meets with the level of demand the companies and their investors are expecting.

By Olivier Discazeaux