Robotics, from Farm to Fork. Interview with Rick van de Zedde

picture by © Gea Hogeveen 

Robotics provides major opportunities to produce food more-efficiently by exploiting new, robust crops in a more-sustainable way”, says Rick van de Zedde, scientist and business developer Phenomics and Automation at Wageningen University & Research. At F&A Next 2018, he will discuss the latest trends and insights into agrifood robotics.

 What is Wageningen University & Research doing in the field of robotics?

“We are one of the few organizations in the world investigating the added value of robotics and artificial intelligence throughout the agrifood production chain – from breeding to quality control and packaging: what applications to invest in, what is the return on investment, and when, and what is the social impact of replacing manual work by robotics. We also develop, together with industry, dedicated software for specific applications. Recent developments include an automated plant sampler that collects pieces of leaves from young plants, working faster and more accurately than any human can. https://www.iso-group.nl/. And an autonomous weeding system that allows farmers to spend their valuable time on other urgent tasks.”

What is the Agrofood Robotics initiative?

“In Agrofood Robotics www.agrofoodrobotics.nl we have brought together all of Wageningen University and Research’s knowledge, expertise and technology about robotics. We can now offer companies a service portfolio that is broader than ever before, supporting them to maximise the new opportunities arising from this technology.”

Under the flag of Agrofood Robotics we have set up an extensive research program – 65 researchers to date – in which we develop, together with industry, state-of-the-art solutions for the most urgent issues in agrifood. For several themes we have established public-private partnerships, such as the use of drones in agriculture. We are also working towards the application of augmented reality (AR) in horticulture. An exciting example here are AR Glasses: these will provide a real-time impression of the condition of a crop or the ripeness of fruits, or provide workers with detailed product information – by simply looking at a plant – making specialized training unnecessary. Our research collaborations include larger companies with the power to invest, and also start-ups who do not have such financial power but are overflowing with ideas and enthusiasm.”

Could you mention some projects you are proud of?

“In creating the precision agriculture tool, Akkerweb, we have organized, together with Agrifirm, an open and user-friendly platform where farmers can collect and examine data about their fields and crops. They can order drone pictures of their crops and get advice on the best moment for fertilization. Farmers pay an annual fee for membership, and then per transaction. The platform is also accessible to start-up companies, for example to test or introduce new apps. https://akkerweb.eu/

Another example is the EU-funded Sweeper project (2015-2018), coordinated by Wageningen Plant Research and with over seven partners from the Netherlands, Sweden, Israel and Belgium. The project will deliver a commercially viable prototype of the first robot for the harvesting of bell peppers, strengthening the market position of the European horticultural sector, which is currently suffering from a shortage of experienced workers. We expect the technology and insights from this project to be transferable to other crops in the near future. http://www.sweeper-robot.eu/

Also, during 2014-2015 in the public-private partnership (PPP) project Program Precision Technology Horticulture, FT-NON Delft, Wageningen University & Research and the partners developed a multi-camera system which accurately determines the orientation of white cabbage, using WUR computer-vision software to test and evaluate the data from the cameras. This measurement concept ultimately formed the basis for the camera module in the the CoreTakr Machine, the iceberg lettuce de-coring machine, launched by FT-NON in July 2016. (see: http://www.coretakr.com/).

A very recently announced new project, NPEC, is funded by the Dutch Science Organization (NWO). Called the Netherlands Plant Eco-Phenotyping Centre (NPEC) it will develop a large-scale phenotyping research facility at two locations, the Wageningen Campus and Utrecht University. We invite potential partners and suppliers to brainstorm together with Wageningen University & Research. The aim is to explore the possibilities around high-tech tools for automated measurements and data collection on how different plant varieties grow in a set of varying environmental conditions, in climate rooms, greenhouses and in fields. www.wur.eu/npec

How do you expect the field of agrifood robotics will develop?

“At Wageningen University & Research we believe robotics provides major opportunities to enable farmers, growers and food producers to be more-efficient and run improve their business practices while becoming more sustainable and reducing post-harvest losses – essential if we are to meet the growing global demand for high-quality food and nutrition with an increasingly unpredictable climate.

 

Developments follow each other at high speed. We are better able to monitor, control and collect data, and predict the growth of crops. This enables the agrifood sector to design the supply chain in such a way that they gain maximum revenue with minimum losses. We have high expectations of deep learning, aka the training of these systems. It works in the same way as a self-driving car. The more data you enter into the system, the more accurate it becomes. There are already several research projects exploring the power of this new artificial intelligence toolkit, such as automated recognition of bell peppers for harvesting, and non-destructive firmness determination of pears using Near Infra-Red Spectroscopy (NIRS). We expect to see many more projects in the coming years.”

What opportunities do you see for start-ups and other companies specializing in agrifood robotics?

“The Market for agrifood robotics is growing rapidly and is expected to become the key domain for robotics R&D and sales in the upcoming years. And there is growing demand among farmers and gardeners for smart, user-friendly applications that provide them with combinations and interpretations of data that they cannot easily obtain themselves. The essence is to find specific applications that exploit these new hardware developments, and that precisely fit the needs of companies in the agrifood domain. This requires a broad range of knowledge and understanding; by establishing multi-disciplinary teams and consortia, I am confident we will succeed.“

 And what challenges do you see?

“Self-learning systems are hot among start-ups. But training such systems requires significant volumes of data, which they often do not have. We could help them with data collection. Via the Greenhouse Challenge, where companies are challenged to cultivate cucumbers with the support of artificial intelligence. We deliver data and sensors to participants that allow them to demonstrate the added value of their innovations.”

What would be your advice to start-ups?

“Define your target group well, and get to know them: what issues and questions are they facing, what preferences, ideas and perhaps also prejudices do they have? Use this information to construct a logical, solid story. You might have a marvelous idea, but in the end you’ll need to satisfy market needs in order to be successful.”