Home PublicationsData Innovators 5 Q’s for Jorge Sáez Gómez, Director of Data Science of Connecterra

5 Q’s for Jorge Sáez Gómez, Director of Data Science of Connecterra

by Eline Chivot
Jorge Sáez Gómez

The Center for Data Innovation spoke with Jorge Sáez Gómez, director of data science and founding member of Connecterra, a company based in Amsterdam, the Netherlands, that uses AI to track the health and fertility of dairy cows. Gómez discussed how Connecterra’s AI platform makes farming more efficient and supports an industry exposed to many challenges.

Eline Chivot: I have read that Connecterra’s Intelligent Dairy Farmer’s Assistant (IDA) would be a “Siri or Alexa for farmers.” What was the inspiration for this idea? In an industry that has gone through many transformations, is it trying to solve old or new problems?

Jorge Sáez Gómez: The initial inspiration came from one of our founders, who used to live next to a farm and has a background in technology. From that moment, it became increasingly clear that there is a gap in agricultural efficiency that can be filled by the application of cutting-edge technology: sensor networks, cloud services, and data analysis. In particular, we saw that artificial intelligence could play an essential role in this, by effectively offering something actionable and to-the-point, which a dump of raw sensor data alone cannot do.

I believe the problems we are trying to solve are not new, but some of the motivators for doing so are. The fundamental problem is about making agriculture more efficient which, for the farmer, has a clear economic motivation: a larger profit margin. For the consumer though, it is increasingly more about animal welfare and fighting climate change. The good thing about this scenario is that, by making farming a bit more efficient, we reap benefits on these three fronts.

Chivot: How does IDA, your AI platform, track cows’ health? What data does it collect and use? 

Gómez: Our platform is an end-to-end solution, and it starts with a sensor that sits around the neck of the cow. With it, we track her movement, and we use artificial intelligence to translate it into her actual behavior: eating, ruminating, walking, and so on. Our AI platform then further analyzes the trends present in these behaviors to come up with actionable Insights, which are often related to the cows’ health: Cows are animals of habit, and if what they do throughout the day deviates significantly from what they usually do, that is often a bad sign.

Some farmers have milking machines that automate the collection of milk and, if they do, we can integrate our platform with them to get additional information and further refine the features we can offer. For example, by having information about how much milk each cow gave every day, we can combine this with the cow’s behavioral data that we gather to generate a “Cow Ranking” for the customer.

Chivot: What are some of the insights you provide to farmers, and how does that help them?

Gómez: A set of insights we provide to our customers has to do with early sickness detection. Detecting an illness early enough and taking appropriate action increases the welfare of the animal and lowers mortality rates. Further even: A study we conducted in collaboration with a research institution has shown that cows that use IDA need 50 percent less antibiotics on average than the ones that do not use it. This is excellent news in the context of increasing concerns about antibiotic resistance.

Another feature that many of our customers find important is our Estrus Insight. Estrus is the period of many female mammals’ sexual receptivity and, in the case of cows, it happens once every three weeks. This is the moment where the farmer should inseminate the cow, but the window for doing so successfully is only a few hours long. Since the only cow that gives milk is the one that has recently calved, missing one of these windows means the farmer will eventually lose an income equivalent to three weeks of milk. By using IDA, many of our customers have improved their fertility metrics significantly.

As a final example, we also provide many of our customers with a Cow Ranking feature, which orders cows by how efficient they are in transforming feed into milk, with the idea that the farmer will selectively breed only the cows with the best genetics. Approximately 80 percent of the costs of running a farm come from the feed that cows eat, so optimizing this is critical for the farmer because of the economic benefit, but also for the environment, since we have no alternative but to make agriculture more sustainable if we want to fight climate change.

Chivot: The farming community might be perceived as one that lags in terms of technology adoption and connectivity. What does your experience reveal, working with farmers?

Gómez: I think the number one reveal would be the fact that it is very easy for technology to get in the way of the farmer if not designed and tested out properly. For instance, there are many systems out there that simply spit out a lot of data in the form of endless spreadsheets or graphs. Many of them are not useful for the farmer, simply because they do not have the time to look into so much data, nor is it immediately obvious what the action should be from it, if any at all.

The second reveal would be that, even if you have a smart product that works for some people, that does not necessarily mean it will work for everybody: Different farming practices, cow breeds, and physical environments make this challenging. This adaptation process is an integral part of the work that we do. Besides, farmers need to see tangible benefits before trusting a new solution, so even when you have a product that is proven to work well, there has to be ongoing communication with the farmer, especially when that initial trust is on the works.

Chivot: Your product is now applied to cows, and used by dairy farmers. Could this kind of technology be implemented for other animals, and in other industries?

Gómez: Indeed, it could; the obvious application would be to use it for other ruminant animals. Since the dairy industry is so big though, we also see a clear benefit in taking the same technology and applying it to everything that happens around the farm. Before the farm itself you have the feed that the cows eat: Could we track where it came from? Could we gather data about it as it grows in the soil? How does feed composition affect milk yield? After the farm comes distribution: Could we optimize delivery schedules, so that milk does not spoil? Could we make the final dairy products more traceable for the customer? We believe that, by working in this direction, we would get a better understanding of this industry, and the quality of the insights we would come up with would be deeper and more interesting than what we actually have access to.

You may also like

Show Buttons
Hide Buttons