July 10, 2019

By Philippe de Lapérouse and Mark Zavodnyik, HighQuest Consulting

Confronted with a shrinking and more expensive labor force, production agriculture (including row crops, permanent and specialty crops, and indoor agriculture) is seeking out and adopting the use of robotics to address not only a shortage in qualified labor, but to improve the timeliness and effectiveness of decision-making based on massive amounts of data.

Combining the use of robots with recent developments in artificial intelligence, growers can now mine massive volumes of data captured in the field to make precise, laser-like interventions in the field, resulting in reduced applications of inputs as well as supporting sustainability initiatives. The global market for ag robotics, estimated by industry analysts to be approximately $3 billion, is expected to increase to $12 to13 billion over the next seven years. The list of companies to watch presented here is not intended to be exhaustive but rather representative of new technologies that are already being used in the field to fundamentally change agricultural practices. 

HARVESTING AND PICKING

Abundant Robotics
Hayward, California
Crops: Apples
Apple picking system co-developed with Washington state apple growers.

The robot uses AI technology to identify ripe from unripe apples. Once identified, the robotic hand uses a vacuum system to pick the apple from the tree and deposit it into a bin.

Agrobot
La Palma del Condado, Andalucia, Spain
Crops: Strawberries
Multi-arm robotic harvesters using AI for strawberry picking.

Robotic Harvesters use AI to determine ripeness and up to 24 arms (each with a camera) to pick by the stem without contacting the fruit.

Clearpath Robotics
Kitchener, Ontario, Canada
Crops: Apples
Developing a variety of robots for use across numerous agricultural applications.

Clearpath robots use automation technology with a mounting platform which can be customized with a variety of sensors, manipulators, etc., to fit specific needs of growers.

Dogtooth Technologies
Royston, Barnsley, United Kingdom
Crops: Berries
Automated strawberry picker with capability to identify, locate, and pick ripe fruit.

The robots use “state-of-the-art computer vision and machine learning software to achieve visual acuity, dexterity, and picking speed,” according to the company.

Energid Industries
Cambridge, Massachusetts
Crops: Citrus fruits
Energid has developed a robotic citrus fruit harvesting system offering significant cost and labor savings to growers.

According to Energid, the robot uses “multiple low-cost picking mechanisms organized into a grid. The extending parts of the picking mechanisms have no actuators and no sensors, making them robust, easy to manufacture, and easy to replace.”

FFRobotics
Benei Dror, HaMerkaz, Israel
Crops: Apples
The FFRobot is a fruit picker with the ability to identify ripe fruit and collect and analyze data.

The robot uses image processing technology and advanced algorithms to distinguish ripe fruit from damaged, unripe, or otherwise unusable fruit. According to the company, the robot will be able to be modified for different types of fruit.

Harvest CROO Robotics
Plant City, Florida
Crops: Strawberries
The Berry 5 is an automatic harvester used for strawberry picking.

According to the company, the harvester uses “multiple robotic components to separate the picking functions of leaf gathering, visual inspection, picking, and then packing. This allows us to build simpler robots, which run faster and operate at the scale needed to support current farm operations.”

Naio Technologies
Toulouse, France
Crops: Leafy greens
The Oz Weeding Robot weeds, hoes and helps farmers during harvest.

The Oz robot is meant for smaller fields and indoor growing environments. It is autonomous and can be controlled with a remote control to schedule, control, and adjust work plans.

Root AI
Somerville, Massachusetts
Crops: Tomatoes
The Virgo 1 is autonomous and meant for indoor farms.  It uses AI technology to identify ripe tomatoes and pick them without bruising the skin.

The robot uses sensors and cameras to navigate the greenhouse and identify ripe tomatoes. Robotic grippers have the capability to reach into vines and only pick ripe fruit. The arm is washable, making it easer for indoor farms to manage disease.

Tortuga Agricultural Technologies
Denver, Colorado
Crops: Strawberries
Developing robots for harvesting strawberries in controlled growing environments.

Robots use robotics, machine learning and AI to pick, pack, and trim strawberries in a variety of indoor growing applications.

SORTING AND PACKING

Robotics Plus
Tauranga, New Zealand
Crops: Apples
The Robotic Apple Packing Cell sorts apples into trays for packhouses around the world.

The Robotic Apple Packing Cell includes a robot with the ability to pick and place the apples into trays, as well as the capability to orient the apples in the correct direction.

SPRAYING, PRUNING, MOWING, AND SEEDING

Agco Corporation (Fendt)
Marktoberdorf, Germany
Crops: Corn
Fendt’s Xaver system is a series of 6-12 small robots which use swarm technology to sow, control weeds, collect data, and monitor corn planting.

According to the company, “Fewer sensors, robust control units, and a clear hardware structure make each individual Xaver robot extremely reliable and productive. At the same time, the use of a large number of small, identical robots operating in a swarm enables smooth running of the job, even in the event of the failure of a single unit.”

Augmenta
Athens, Greece
Crops: Row crops
Field Analyzer plug-n-play device mounts on tractors to help farmers more accurately fertilize.

The Field Analyzer uses AI technology to compile 4K video using hyperspectral computer vision and crop-focused intelligence to apply the precise amount of fertilizer.

Autonomous Solutions
Mendon, Utah
Crops: Broad acre crops
Autonomous tractor vehicles capable of operating 24 hours per day.

According to the company, “Each unmanned farm vehicle is equipped with a series of hardware and software components that permit a user to toggle between manual and robotic control. The components work with the existing by-wire, mechanical, or hydraulic control system and link vehicles to a central command station, allowing a single operator to simultaneously manage multiple vehicles throughout a farm operation.”

Bear Flag Robotics
Sunnyvale, California
Crops: Broad acre
Row crops
Orchards
Vineyards
Self-driving technology for farm tractors.

Bear Flag’s technology uses perception sensors and robotic actuators to navigate fields and perform a variety of common farm tasks.

Blue River Technology
Sunnyvale, California
Crops: Cotton
Soybeans
The See & Spray uses smart technology to more precisely apply herbicides.

The See & Spray uses computer vision and artificial intelligence to differentiate between plant and weed and apply herbicide only to the weed.

Deepfield Robotics
Ludwigsburg, Germany
Crops: Sugar beets
The BoniRob robot navigates fields, identifies and smashes weeds.

Rather than picking weeds, the robot uses a camera to identify them and stamp them down into the ground.

ecoRobotix
Yverdon-les-Bains,Switzerland
Crops: Row crops
100% autonomous robot uses solar power to precisely apply herbicide.

Rather than human operation, the robot uses a camera, GPS and sensors to navigate and identify weeds. Robotic arms apply herbicide. Solar powered, the robot can run up to 12 hours at a time.

Rowbot Systems
Minneapolis, Minnesota
Crops: Row crops
Self-driving rowbot equipped with sensors and GPS technology capable of spraying fertilizer.

Rowbots are intended to work in teams, navigating a row crop field during high growth season, and spraying fertilizer in precise amounts.

TartanSense
Bangalore, India
Crops: Cotton
Robotics and AI solutions for small farms.

Semi autonomous robots mounted with cameras and pesticide sprayers use AI algorithms to identify weeds versus plants.

Wall-Ye
Macon, France
Crops: Grape vineyards
Mobile vineyard pruning robot.

The robot uses AI and GPS technology to navigate the vineyard, identify key plant features, capture and use data, and prune leaves.

THINNING AND WEEDING

AgMechtronix
Silver City, New Mexico
Crops: Lettuce
Automated row crop thinner

According to the company, “the Agmechtronix Row Crop Thinner (RCT) is capable of thinning row crops such as lettuce using computer machine vision to identify plant locations and a herbicidal spray to eliminate the unwanted plants, replacing the traditional thinning method of hand labor.”

F Poulsen Engineering
Hvalso, Denmark
Crops: lettuce, cabbage, fennel and onions
The robovator is used to control weeds in row crops

According to the company, “The robot  is equipped with a special plant detection camera above each row. It has a mechanical tool which is operated by hydraulic power.”

TerraClear
Bellevue, Washington
Rock clearing
Developing an automated solution to clear rocks from farmland, a process normally involving  significant manual labor.

TerraClear’s system integrates GPS technology, drone capability, machine learning, and robotics to clear rocks from farms – a solution to a labor-intensive problems.

Vision Robotics
San Diego, California
Crops: Lettuce
The VR lettuce thinner uses intelligent spraying technology to identify unwanted plants. The thinner can cover 2 to 3 acres per hour.  

The VR lettuce thinner uses cameras to identify and understand unwanted plants and only direct precise spray at those plants.

NOTE: the below four points were included as sidebars in the printed version.

LABOR SAVING  

Strict enforcement of increasingly stringent U.S. immigration laws is increasing the difficulty of hiring competent and qualified labor in production agriculture. While the national median annual cost of a U.S. farm worker in 2017 was $23,730 ($11.41 per hour), the median cost for farm labor in California, where the adoption of robots is experiencing the fastest growth in the U.S., is $20 per hour. According to a California Farm Bureau Federation survey conducted in 2017, 55 percent of responding growers have experienced labor shortages in recent years. For a majority of growers who are growing high-value crops, such as berries and grapes, which are difficult to harvest mechanically, the labor shortage is particularly challenging.

The adoption of robots as an integral part of farm production provides labor cost savings and generates increased operating efficiencies. For example, the Harvest CROO Computerized Robotic Optimized Obtainer can pick a single strawberry plant in eight seconds and cover eight acres of strawberry fields in a single day, replacing the labor of 30 pickers.

DRONES 

While drone or unmanned aerial vehicle (UAV) technologies are not covered in the 25 companies included on the “Companies to Watch” list, drones are playing a key role in enabling growers to collect more actionable data on a timely basis, facilitating faster and more informed decisions.

AgEagle of Neodesha, Kansas, uses UAV aerial imagery for data collection and to conduct analytics supporting sustainable and precision farming practices. For commercial growers, AgEagle’s drones take thousands of high-resolution images to produce detailed prescription maps, identifying diseases, pests, weather impact, etc.

The use of UAV technology is also taking off in Europe. Unlike other players in the drone imaging space that use multispectral imaging to map farmland, Gamaya, a Swiss developer of drone imaging technology, has developed a lightweight and data efficient 40-band hyperspectral camera combined with artificial intelligence to capture and analyze crop information that growers can use to significantly enhance their decision making. While it is a more expensive technology to use, hyperspectral imaging provides increased insights to growers via the use of multiple light spectrums that capture the reflection off of plants to measure the chemical composition of those plants. This provides insights on the plant’s health and development during the season. Combining this technology with advancements in artificial intelligence, Gamaya believes it can increase the efficiency of capturing data in the field and improving  productivity, while minimizing the environmental impact caused by intervention in the field.

FUNDING

The following are examples of recent investments in ag robotics:

Most notably, Blue River Technology was acquired in 2017 by John Deere for $305 million. Blue River’s See & Spray devices combine machine learning with robots to pinpoint exactly where on a plant the application of herbicides will be most effective, thereby significantly reducing the overall usage of crop inputs.

More recently, the Swiss company EcoRobotix developed a 100 percent autonomous robot powered by solar energy to apply herbicides. The company announced in 2018 that it had raised $10.7 million in Series B funding led by the French ag innovation fund Capagro and BASF’S venture capital group. Rather than relying on operation by a human, the robot uses a camera, GPS, and sensors to navigate and identify weeds to be treated by robotic arms applying the herbicide. Solar powered, the robot can operate autonomously for up to 12 hours at a time.

 SIZE OF FARMING OPERATIONS USING ROBOTICS

The increasing adoption of robotics in agriculture is not just taking place in large-scale row crop operations. It is also being deployed in emerging markets where small scale agriculture is more pervasive.

In India, TartanSense is developing robotics and AI solutions specifically designed for use in small farms. The company’s semi-autonomous robots are mounted with cameras and pesticide sprayers that use AI algorithms to identify weeds to be destroyed versus plants. TartanSense was founded by Jaisimha Rao in 2015 after he observed a lack of data-driven decision making on his own family’s coffee plantation.

In order to support the growing indoor farming sector, Massachusetts-based Root AI has developed a robot called Virgo 1 that uses sensors and cameras to navigate greenhouses to identify ripe tomatoes. The robot’s grippers can reach into vines to identify and pick only ripe fruit. The arm is also washable, enhancing the ability of indoor farms to avoid and manage diseases.

CONCLUSION

While agriculture throughout history has demonstrated that new technologies providing enhanced efficiency and labor productivity will be adopted rapidly, new technologies adopted from adjacent industries such as robots integrated with artificial intelligence promise to accelerate the pace of change, and revolutionize how food and crops are produced in ways we have yet to fully appreciate.

ABOUT THE AUTHORS

Philippe de Lapérouse is a managing director at HighQuest Partners, a leading global strategy advisory and consulting firm. HighQuest advises strategic players operating in and financial investors allocating capital to the global food and agricultural value chains on making informed decisions on strategy and resource allocation. Lapérouse chairs the Global AgInvesting conference series. He can be reached in St. Louis at +1 314.960.1632 or via email at pdelaperouse@highquestpartners.com.

Mark Zavodnyik is project manager for HighQuest Partners where he leads the day-to-day execution of consulting projects, advising clients on strategy and investment decisions across the global agricultural value chain. Previously, he was the lead tropical oils trader at AAK USA with responsibility for all sourcing, trading, and risk management in the United States. Zavodnyik has spoken at industry conferences on the efforts the industry has undertaken to make palm oil more environmentally sustainable. He can be reached at +1.574.274.3099