Trip report from the Carolinas and Georgia

I live in Northern California, was born and raised in Brooklyn, spent childhood summers in Michigan, and lived in Houston for college and my early professional years. I identify with all sides of the Soda-Pop-Coke debate, and hope that I have more than a simplistic, bi-coastal perspective of the US.??

Still, during my recent trip through the Carolinas and Georgia I was very surprised by how much stuff is being built today in the region: chemicals, cars, food, and much more. Driving along the interstates you see factories left and right manufacturing a wide variety of goods. (Also a lot of Waffle Houses, which I love – my hashbrowns are smothered and peppered!) I suppose I knew this as a rational business person, but it’s one thing to read about it and another to see it live.?

Last fall, a few of us started working on an AI startup that we are calling TC Labs for now. Our purpose is to help factories run as efficiently as possible. We use cutting edge AI tools and methods to assimilate all of a facility’s data and use that data to help the people running the factory improve operations and spot problems before they crop up. We can boost profitability right away, mostly by improving energy efficiency. Since factories use about 37% of global energy annually, improving their efficiency can have a major climate impact.

We’ve been working developing our product, and my trip was a chance to get out there and talk to potential customers about what’s going on with manufacturing today. I saw first hand many of the things that our team has been discussing. Sharing here, as I thought my LinkedIn network might find it interesting.

The data environment is bad, but not a disaster. The data problem in factories is similar to what I saw in other industries when working with Analytics products at Google. There’s a lot of data and much of it is high quality, but it is in silos. It may be structured but with different details, making it very hard to compare apples with apples. There are big gaps. Some aspects of the process are data rich, others are data deserts. In one place I visited the best available data about one key step in their process could be found in hand written notes. He showed me the notebook!?

Creating a holistic view of what’s going on in a factory requires solving this problem.??

The teams running these factories are smart and disciplined, but their data analytics expertise doesn’t come close to matching their manufacturing expertise. They reminded me of the engineering teams I used to work with at Google to bring them into the data analytics age.They were building some of Google’s core products, but building software as they always had. They had to be shown how to use analytics to get better.

The factory leaders I talked to were just like that. They were doing things the way they always have, because no one had shown them a data analytics tool that would actually be useful. When I demoed ours, they lit up. They are eager to embrace analytical expert systems that make their work better and easier.??

However, these leaders and teams have built up a healthy skepticism of technology. They have lived through a litany of rules-based tech solutions that don’t easily conform to the realities of how their plant operates. These factories are so complex and (duh!) physical; there’s a big gap between the ideal world and the real world. The people on the floor need to trust that any expert system they use is based on a model of how the plant is actually operating right now.??

I came home from the trip with an even greater appreciation for the people who run our factories. They work in a far more dynamic environment than ever. Factors change all the time: input materials, pricing, product mix, and so on. When they do, it’s not like rewriting a few lines of code. They have to go out on the floor and make physical changes. I was in awe getting to see them in action.

One more thing I learned: North Carolina BBQ is different from South Carolina BBQ, and they are serious about it. The north uses a mustard-based sauce, while the south favors a vinegar base. Or maybe it’s the other way around. It was all delicious!

Not as great as Waffle House, though!?