Employee Spotlight: A day in the life of an agtech data scientist

I’ll be honest, it’s hard to keep my attention. I’ve never been someone who liked working on the same problem everyday for long periods of time, I learned this while getting a PhD. So when I got a call back from advanced.farm, I bought in; hook, line and sinker. You want to work out of a barn developing software that will run on robots? Check. Want to develop interfaces for developers to see what the system was seeing so we can more quickly diagnose and correct problems? Yes. Want to sometimes ride on the side of a prototype, auto-driving tractor, kicking un-trimmed branches out of the way so that testing can continue? Sure do, let’s automate specialty crop farming! 

More than crunching data on a spreadsheet

I’m a data scientist in title, but that doesn’t tell the whole story, and not by a longshot. Yes, a good number of my days are spent visualizing data, finding trends, and providing insight to the various engineering and business teams. Other days, however, and aside from kicking branches, I’m building APIs for reporting errors on harvesters and deploying code across our fleet, crawling behind a strawberry harvester watching every pick attempt and getting very upset when it misses one I think it should have picked, or watching videos generated by our robots “eyes” and doing my best to figure out why it might not have done what we expected. Heck, a good portion of my time at advanced.farm has been spent breaking things on purpose just to make sure we can quickly, accurately, and concisely report when something has broken, or when a piece of software has failed. I build dashboards to monitor our performance in the field, and I break 3D molden fingers just to see how the sensors behave. It’s not at all what I thought of when applying for data scientist positions.

Farm robotics is an emerging field with no shortage of opportunity

There is no shortage of problems to solve in robotic farming, and this makes it the ideal environment for someone like me. If there is a problem I am interested in, odds are it needs someone to take ownership of solving it. It’s challenging and fulfilling in a way I never would have imagined for my first job out of graduate school after taking a major left turn and completely changing fields. Most importantly, it’s fun. This may be a cliche, but I genuinely feel like we are building something impactful as a team and I wouldn’t have it any other way.

By: John Bridstrup

Interdisciplinary scientist with a background in classical and statistical physics, computer simulation, machine learning, software development, finite-element modeling and nanotechnology. Skilled in Python, PyTorch, Sci-Kit Learn, Data Analysis, MATLAB, and other various soft and hard engineering processes. Strong independent and collaborative problem-solving abilities, with a PhD in Physics from Drexel University.