xBerry Talks – The Future of Agriculture with Chris Parjaszewski and Marek Kraft

When an interesting and exciting subject meets experts from the field who have some thoughts and knowledge to share – here come xBerry Talks. In this edition of the xBerry Talks series, we discuss the future of agriculture and technology’s influence on the farming industry with Chris Parjaszewski, CEO of sky gate, and Marek Kraft, Ph.D. of Institute of Robotics and Machine Intelligence from Poznan University of Technology. If you’re not familiar with the topic yet, check out our last blog post about Smart Farming and Automation below this link.

The Farming Crisis and the Search for Solutions

The recent geopolitical events in the world and the advancing climate crisis have had an impact on agriculture. The war in Ukraine exacerbated the food crisis, damaging food supply chains. At the same time, rising temperatures resulted in the worst droughts in Europe. Poor weather conditions affected the harvest and less food than expected ended up on the market. Only in Poland, during this year’s spring, considerable water shortages for arable crops, ranging from -50 to -179 mm, were noted over a large area of ​​the country. The prevailing weather conditions during this spring, due to unusually generous sunshine and high wind speeds, resulted in a loss of crops due to water shortage in 14 voivodeships. Can technology provide an answer to these worsening problems in the farming sector?

“Widely understood precision agriculture has great potential for improving the efficiency of food production and lowering the use of resources,” says Doctor Marek Kraft. “The solutions available under this umbrella term include monitoring and decision-making tools enabling more efficient resource allocation, be it water, fertilizers, or plant protection products. Continuous monitoring, in turn, facilitates faster reactions to events like pest infestation or nutrient deficiency. This way, we can achieve more with less – improve the yield while decreasing the use of water or pesticides. Actions like soil preparation, seeding, on-site inspection, weeding, spraying, and harvesting can be performed by increasingly autonomous machines at exactly the right time, which is calculated based on measurements and analytic functions. Nevertheless, regardless of the technological advancements, I think that without any change in the structure of food consumption towards a more plant-based diet, we still might be facing serious sustainability issues in the long run.”

The progressing crisis in the agricultural sector is not only about shortages of primary substances such as water or fertilizers necessary for proper and effective farm management. Chris Parjaszewski, CEO of sky gate, notices several problems in the industry that negatively affect the situation of both farmers and food consumers:

“We can see in every industry – including agriculture – that improvements start with numbers and data – in the sense it is hard to improve a process if it is not dimensioned and not reflected in reliable numbers. Let’s take as an example the case of apple production in Poland. The problem here seems not to be the product itself but the logistics. Over 50% of apples produced in our country lose their value during transport. The same fruit, harvested from the tree, for which the customer could pay even PLN 5-10 per kilogram in a boutique shop, is sold for 0.6 PLN as an industrial apple for juices due to ineffective logistics. This data is, of course,  averaged for the country as a whole, but my point is the implementation of innovations begins with this kind of analysis,”  says Chris.

In the field – Robot vs. Human

While observing the development of technology, we like to fantasize about replacing human work with more and more modern machines. The spread of AI, autonomous robots, IoT, and drones allows us to predict that the modernization of the agricultural industry is already doing well, and we can expect more and more innovations. In the USA, revenue from the global agricultural robot market is expected to register a CAGR of 34.4% over the forecast period. Turnover is projected to increase from USD 6.94 billion in 2021 to USD 99.30 billion in 2030. Will these numbers translate into practice, and can robots replace human workers in the agricultural industry?

“Robots are mostly thought of as machines that take over tiresome, repetitive, or otherwise tedious tasks. There is no shortage of such work in farming, so if intelligent machines achieve a level of capability and autonomy that’s high enough to perform a job that isn’t particularly attractive, they will eventually be used. I don’t think they will replace the people completely but most likely in the near future they will gradually take over the most tedious tasks” says Doctor Marek Kraft. What does this mean in practice? 

“We can run the same farm in the growing season with 20 employees, and in the off-season with 4 employees and machines worth 5 million PLN. Then we can buy machines for another 5-10 million PLN and reduce the number of employees in the growing season by half – here, of course, you have to deal with a piece of paper and a calculator and calmly estimate potential profits and losses. Theoretically, we can automate EVERYTHING, but not everything pays off. We’re certainly witnessing increased labor costs in Poland and other countries so, relatively speaking, machines are a better solution. But on the other hand, it may be that the prices of machines and operating them will go up – I think it will never be black and white,” answers Chris Parjaszewski.

According to the Emergen Research report on Agricultural Robots Market, a reduction in the workforce in the agriculture sector, deployment of agricultural robots and the possibilities offered by working with different types of innovative machines, as well as a shift to more advanced farming techniques and approaches are trends that are expected to continue and, as a result, support revenue growth in this market.

“Self-driving heavy machinery for plowing or harvesting is already being employed. Machines requiring more advanced perception and control, such as weeders or sprayers, are also becoming reality. I also believe that agriculture will become more data-driven. Sensing with IoT-like solutions, unmanned aerial vehicles, or robots provides a lot of useful information. Together with applications that make sense of this data, they are another piece of the puzzle, and their use is on the rise,” comments Doctor Marek Kraft. “Technology has reached a stage where it is actually useful in many ways and is constantly being developed. Field deployment is always hard and relatively small batch production results in relatively high cost for market-ready solutions, but as they become more popular and commonplace, I think the prices will drop eventually. It’s a really exciting field open to new deployments, ideas and developments, and the incoming innovations may have a great impact on our future lives.”

Theory in practice

We know the challenges faced by the agricultural sector. We are also well-acquainted with the technological tools available for us to work together and modernize this sector while also preventing the multiplication of problems and crises. 

What projects and technologies are already supporting the transformation of agriculture into agritech? What is going on in farms and companies specializing in solutions tailored to the needs of agriculture?

“As sky gate, we took part in two projects focused on agritech – one of them involved observing fruit development using cameras and Computer Vision in a continuous mode. The second one was about detecting insect attacks using an innovative system also connected with cameras,” says Chris Parjaszewski. “In the first case, trying out the model was very expensive, and the number of cameras was high, so the project turned out to be unprofitable. In the second case, an insect trap could cost about PLN 2,000-3,000 gross and was a real saving for the owner of a several-hectare field. Our farmer did not have to carry inspections out several times a day, and the system counted the number of insects every 4-6 hours. This project brought two benefits: the time saving and greater accuracy and systematicity, thanks to which the farmer could introduce preventative measures much faster and more accurately. What have these two projects taught us? Simple solutions are better because they are easier to implement and facilitate improvements in the process.”

Doctor Marek Kraft also shared his experiences:

“I’ve had the opportunity to work on a range of precision agriculture projects, including automated weed detection, an autonomous hemp plant treatment robot, and a range of applications involving crop and orchard inspection using unmanned aerial vehicles. I think all these projects generate some added value, either by directly making the farmer’s job easier or by generating useful insights and analyses facilitating better decision-making. The key outcomes are usually a reduction in labor or better yields. Sometimes both.”

Agriculture is one of the oldest industries that remains crucial to humanity to this day. To successfully overcome the problems that have hit it and modernize our processes, we need high-level solutions ideally suited to the needs of farmers. Today we know that technological transformation determines the success of this task – and thus the future and quality of life for future generations.

“People working in agriculture are trustful and open-minded. But to achieve the best level of cooperation, you need to have reasonable ideas and discuss them a lot – I mean more like asking questions and not trying to pretend that you know better than the farmer. They are very wise, talented, and hardworking people – so the best strategy here is to listen carefully and understand each other,” summarizes Chris.

Chris Parjaszewski
CEO
sky gate

A graduate of the Faculty of Mathematics and Computer Science at the University of Wrocław, an entrepreneur for 17 years, a father of three children, and a business angel for 5 years. Fascinated by young enterprises at an early stage of development.

The founder and CEO of sky gate – a software house with development teams in Europe and the United States – he gained experience mainly working with the best accelerators such as Y Combinator, 500 Startups, TechStars, Plug and Play or Launch House. 10 of his clients have achieved the status of a centaur or unicorn – that is, respectively, companies valued above $ 100 million and $ 1 billion for several years, and therefore with a very high growth rate, mainly young technology companies based in San Francisco. Investor in sky vision.

Marek Kraft, Ph.D
Institute of Robotics and Machine Intelligence
Poznan University of Technology

Marek Kraft is a long-term employee of the Faculty of Control, Robotics and Electrical Engineering of the Poznań University of Technology. In 2005 he graduated from the same university with a Master’s degree in Automation and Robotics (specialization: Automation). In 2013 he defended his doctoral dissertation, obtaining a doctoral degree in the field of Automation and Robotics. He is also a graduate of post-graduate studies “R&D Project Manager” at the WSB University in Poznań. For over 10 years, he has been working in the field of image processing, machine learning and robotics, and in engineering practice, he has also been involved in designing embedded and measurement systems. His experience in R&D projects includes cooperation with startups and companies such as bin.e, Lerta, Solaris, HoloSurgical, Inteneural, Kardiolytics, Surgalign and Inwebit. He is the author of over 50 scientific publications and a member of the international organizations IEEE (Signal Processing Society) and ACM.

Areas of competence:

• broadly understood processing of images and video streams, with particular emphasis on machine learning methods, including methods based on deep convolutional neural networks (deep learning):
• recognition, detection and tracking of objects in images and video,
• binary and semantic segmentation,
• medical image processing,
• practical use of the above algorithms and methods for vision-based navigation, medical diagnostics, autonomous vehicles, safety and monitoring,
• implementation of the above-mentioned methods in embedded architectures.