CTO interview: Jane Cummings, the physics genius who went from CERN to AgriTech
This week on Tech Captains I interviewed a CTO from a very interesting industry, that I am less interacting with: AgriTech. Agriculture is no longer only about cows, crops, and heavy machinery, but also about advanced software technology!
Read how a talented tech person can find solutions to problems in any industry. Jane Cummings, who has a PhD in Physics and worked at the CERN, is now helping dairy farmers.
You have an impressive academic background with a PhD from Yale. You then started at Cainthus as a Data Scientist and became a CTO over the years. What was the impact of your academic background on your professional success?
Interesting question. I studied maths and physics for my undergrad, then did a PhD in physics at Yale. For my PhD research, I worked at the Large Hadron Collider in Switzerland.
I’m excited by working on really challenging problems and problems with significant uncertainty that have never been solved before. That started with my PhD at the LHC and that’s where I developed a lot of the skills that are needed to be successful in enduring through technical complexity and uncertainty.
I could bring some of my previous experience of “exploring the unknown” to the innovative problems that we are solving with computer vision on dairy farms at Cainthus.
I’m more used to interviewing CTOs of companies in the FinTech area, or various apps. Your field of work is unusual for me. You solve different problems in farms. Can you please tell me more about that?
We’re developing a monitoring product for dairy farms that enables farmers to be more efficient, profitable, transparent, and support animal welfare. We have developed a product based on computer vision: we install cameras on farms and our software provides intelligent observation of the farm 24/7.
We’re converting observations into insights for the farmers that will help them to improve their productivity and efficiency. It is based on the idea that if you can make measurements, you can manage your system better.
We measure things on the farm that are not measurable with existing technology and processes, especially not at the scale that we can provide.
Can you provide some examples, please?
Sure. An important part of dairy production is to provide feed to cows throughout the day. Dairy cows need to eat in order to produce milk. Feed is one of the most significant costs for the farm. It is important for dairy farms to optimize around the total amount of feed that is provided each day to maintain high productivity while also preventing waste. We provide observation of the amount of feed throughout the day and we help farmers to manage their feeding operations efficiently by monitoring whether they are providing the optimal amount of feed, if it is enough for all cows, while also making sure there is minimal waste on a daily basis.
Technically it sounds pretty complicated. Especially for visual recognition in farms that are often dark and messy. What is your tech stack?
One of the early challenges that we faced is the variety in farm environments. When developing a computer vision solution, you need to be aware of variations in the data that you’ll encounter — different lighting conditions, physical changes in the scene and environment, obstructions, seasonal changes, and various different edge cases. What you’d really like to have when developing computer vision solutions is a constrained environment with a high degree of similarity and predictability between different environments. Dairy farms are similar enough that they are a great use case for computer vision solutions, but we did have challenges in the early days when we only had data from a few farms in R&D. Now we have built up a more comprehensive dataset that represents the range of variation in the environment and this has enabled the generalization of our technology.
In terms of our tech stack, we have a hybrid solution with cameras, networking, and edge computing devices on the farm and our software running on the edge and cloud. Our customers have access to our insights in the form of dashboards and analytics available through a web application. We leverage the cloud as much as possible to increase our pace of development. All of our software is written in Python, which has also been very important for us both for the speed of development as well as cohesiveness as a team. We have the technical complexity combining hardware, software, and machine learning, and it’s important that all of our engineers can work together with a common language.
Can you describe the tech ecosystem you see in Ireland?
The Irish tech ecosystem is very active and it is a great environment for us to be based. There are a lot of startups, meet-up groups, conferences, and opportunities to meet engineers working on interesting problems in technology. We work in a co-working space called DogPatch Labs that is home to many other startups and it is an especially good tech community.
What advice would you give to younger techies who don’t know what direction to go?
I think it depends from person to person, however the best advice I’d give is to be curious and try different things until you find what excites you. If you’re not sure what direction to go, it can be helpful to reach out to people in different industries, roles, and different types of companies and ask them about what they do. If you think that you might be interested in working at a startup, I recommend giving it a try. It can be a fantastic learning experience and a great environment for growth.
What I like about working in a startup environment is the opportunity to take on big challenges, working on problems that have never been solved before, the fast pace, and working alongside others that share the same values and passion for solving these problems.
If you want to connect with Jane, click here.
To learn more about Cainthus, visit their website: cainthus.com
If you’re a techie working on something exciting or you simply want to have a chat, get in touch with me. I’m currently CTO at Kolleno.com
FROM THE AUTHOR
>> 5 reasons your invoices are being paid late
