You are a veterinary and IT expert. Why is this combination special?
To be veterinary and IT expert at the same time is being the representative of a profession which doesn’t exist yet. I learned veterinary medicine and worked many years in IT positions. As I see, there is a huge understanding gap between people with professional knowledge, and IT knowledge. They don’t understand each other’s problems, opportunities, or even thinking. That’s why there are so few satisfying solutions on the long term.
This is particularly true for agriculture, which is quite often ruled by personal experience and anticipation. However, data science can process, understand, filter the information, and come to real business conclusions. I can bring these two areas together, as the business requirement, the analysis of the problem, and the IT solution are all coming together in my brain.
With the spread of agricultural systems and IoT, the data streams in agriculture are growing exponentially, and this means a gold mine for someone like me.
Data is the new oil, which catalyzes the economy, and within that, agriculture.
What is the professional background you brought to TE-FOOD?
I graduated summa cum laude at the Hungarian University of Veterinary Science in 1981. Some years later I met computing in the form of Commodore 64, and it became the second love.
At that time, it seemed so fresh and strong, that I started to work in a team to develop software games. Hungarian game development in the 80s was quite strong, with famous titles like Eureka!, Impossible Mission II, or Ecco the Dolphin were developed here.
Later I worked at several multinational companies like General Electrics, Raiffeisen and Ernst & Young, many times on projects which combined agriculture and IT.
What did you like in TE-FOOD which convinced you to join the team?
TE-FOOD and me were an instant match, as both of us bring hypes to the ground in agriculture. We connect two areas, which — at least today — are far away from each other. TE-FOOD introduces new technologies to food companies, my part is to show them the reality through the lens of data.
Being pioneers in this area is extremely uplifting and difficult at the same time. We have to break through walls in business, technical and human habitual sense as well.
It’s not possible to do it without commitment, passion and determination. You can’t change these things doing a traditional 8 hours job. It needs to become a hobby, and eventually, your life.
What was your main focus at TE-FOOD during the last months?
Decision making in agriculture is mostly based on experience and anticipation: “I learned it from my parents, and polished my knowledge with decades of experience” they say. Many of them think agriculture is unpredictable, however, scientific approach can bring predictability.
Working on the TE-FOOD Deep Data Analysis module with my colleague, Katalin Vereczkey, sometimes we feel like those who invented microscope. Before microscope, people thought illnesses are caused by demons, or curses. The microscope helped to find the pathogens, which led to solutions.
Data science combined with agricultural knowledge and experience help us to find the “pathogens” and the solutions in agricultural productivity. But we have to know the limitations of farmers, so we don’t show them complex diagrams, the software generates customized, understandable sentences with practical instructions.
During our tests, following the suggestions shown a profit increase of more than 10% after the first year.
It seems like a lot, but today, in many occasions agricultural production has just 40% productivity. We can increase this 10+ percentages from this undiscovered potential. Our solution doesn’t require extra investment, we just show them how they can repeat formerly achieved good results.
The Deep Data Analysis module show the factors which increase or decrease their productivity and profitability, and formulates them into professional sentences what they understand.
Certainly, we still have a long road ahead of us. Improvement is an endless process, but this is the beauty in it. Biology and data analysis fit together perfectly, because the structures — which was refined by evolution during millions of years — are stable and reliable, and data analysis can reveal their characteristics, and their interaction with their environment.
Our next tasks are to introduce more statistical calculations to the system to improve our analytics, to decrease the human parametering tasks by implementing machine learning elements, and to reveal more hidden correlations with AI.
What makes the TE-FOOD Deep Data Analysis a unique product? What kind of productivity improvement suggestions will the system give to farmers?
For example, there was a pig fattening farm with a yearly net revenue of 4 million USD. Our rotation based analysis showed, that their production spread was around 120%. On a yearly level it means that there is a 600 000 USD difference between better and less good production groups. Their net revenue was depending on having groups with better productivity and profitability in fattening pig growing to counterweight the groups with worse parameters.
But the factors of this discrepancy were something they didn’t know about. To reveal the discrepancy enabled the software to find and spread the good factors, and eliminate the bad ones.
The system usually calculates with 80–100 factors, and 8–10 key performance indicators (KPIs). We collect data about the animal husbandry processes, foraging, veterinary checks, weather conditions, from everywhere we can gain digital data in the environment of the animals. For us, IoT is a gold mine, and its spread will revolutionize this industry.
Many times, there are less than a dozen factors, which significantly improves or worsens the KPIs. Our final suggestions cover those factors, where the correlation with KPIs are the strongest, and they effect more KPIs. We suggest farmers to run the analysis bi-weekly or monthly to refine the results.
What is the target market of TE-FOOD Deep Data Analysis?
TE-FOOD started this module as a part of the traceability product package, to motivate and incentivize farmers to implement and use it. As in traceability solutions, the biggest cost is on the upstream supply chain (farms), while the biggest benefit on the downstream (retails), this module is meant to give back something for the farmers, to mitigate the disproportion of costs. But I believe in the future it will be a standalone product as well.
The module can be used in the whole production process, for crop production and animal husbandry as well.
Agricultural Deep Data Analysis is useful for farmers, agricultural investors, forage producers, and agricultural consultants. The size of the farm is only important for the level of data volume and availability.
Some of the farmers are worried about their existence, because the system will provide a clear picture about their production problems.
To implement this solution, we have to educate them continuously, and explain that this is a novel way to extend their anticipations and experiences with scientific knowledge and mathematical algorithms.
How do you imagine food supply chains will look like in 10–15 years? What technologies will they use? How will they differ from their today’s processes?
With the current pace of technology, 10–15 years is too far away to predict.
I would say, in 4–5 years time, IoT technologies will spread like wildfire. It will be intelligent and cheap enough to be used in all kinds of farms. Recently on the European Acceleration Summit in Hannover, Huawei specified Agricultural IoT as one of their most dynamically developing IoT device segment.
IoT will replace several manual entries during food traceability with automatic data services. It will drastically improve the reliability of traceability systems and support their penetration. These changes will result in optimizations within the food supply chains, and the authority control will be easier.
Paper based documents will be replaced by blockchain based ledgers, and food supply chains will be much more transparent. As a result, food will be not only more healthy, but cheaper as well.
I also expect process mining (data mining of traceability data) to be a developing data science area, so there is plenty of opportunities left in this field for the next years.
TE-FOOD is the world’s largest publicly accessible, farm-to-table fresh food traceability solution. Started in 2016, it serves 6000+ business customers, and handles 400,000 business transactions each day.
Website: www.te-food.com
Telegram group: https://t.me/tefood
Twitter: https://twitter.com/TE_FOOD
Reddit: https://www.reddit.com/r/TE_FOOD/
