Let’s start from the beginning. How did you get interested in automation?
It is kind of a twisted path. Growing up I was really into computers and software. However I started my career in finance with KPMG in a feeder program for investment banking, and I learned a lot about the fundamental of financial metrics, which was a great foundation for later, but I hated the work.
At that point I started looking around, and it really shook my world when I realised I didn’t know what I wanted to do. At a certain point I told myself “well, strategy consulting sounds like something for people that don’t know what they want to do.” This was one year out of college.
I had a friend who was working at Accenture Strategy in Singapore, I like to travel, and I had a double major in Economics and Chinese because I thought that’s where international business would be: so I had an interview with Accenture, and I ended up going to Singapore to work for them. The person I was working for was phenomenally good. I learned a ton, the work was varied enough to be interesting and fairly technical which was sort of serendipitous.
During the first projects I worked on, I had some exposure to this new thing called Machine to Machine Communication (M2M). That would eventually become IOT. I got very interested in that, and I made it my focus. I began working on those projects for the customers.
But then after two years, you moved back to the U.S.
Yes, when my wife and I found out that we were expecting our first child. I went back with the firm to Seattle, but I was still only working for international clients. I was flying all over the world from Seattle. It was fun, but it was also crazy and tough with a kid: I was in Thailand and Indonesia one week, and the next week I needed to be in Boston and then Amsterdam. I never knew what time zone I was in, and my teams were all over the world. It was not sustainable.
I also was starting to crave more technical work because we were still very much focused on the business side. About that time Peter Carlsson was talking to my old boss in Singapore as he was looking for someone to work on connectivity issues at Tesla. That is how I started working at Tesla building their global connective cars strategy.
It ended up being a super fun job. I would work on the hardware we were going to use, but also on the software components: how do we manage the services and how do we do it securely? And of course, there was a major commercial element too in my job: I’d get to meet with the CEO of AT&T or Telefonica and I learned how to build complex high stakes partnerships to solve technical problems.
It sounds really fun.
It was really fun. I did it for about a year, and then I wanted to do more, take on more responsibilities. I wanted to do more engineering. At the time, I was writing a lot of software in my spare time, but I didn’t get to do that for work. Then I started doing it at night and weekends for problems we had at Tesla. Over time, those things I had built began to be used by more and more people in the company, and systems were starting to depend on them. That’s when I started hiring some engineers, and I ended up building up a small software team.
I worked on connectivity, autopilot… one of the first big data things we worked on was taking medium accuracy GPS data and using machine learning to combine that into very high accuracy maps. My built the first code and infrastructure that programmatically checked those thousands of miles of map data for errors.
Around that time, Elon started talking about automation being the next big thing. I thought would be the most exciting thing to work on. I told everyone I could at Tesla that it was what I wanted to do, and eventually I think they got tired of me asking and so they gave me the chance to built an engineering team whose goal was to automate Tesla’s supply chain, material flow, and warehouses, which was an incredibly interesting software and sensing problem.
Why do you think automation is crucial for the industry?
When people ask you “what is Northvolt’s thing? What do you guys do?” I don’t think that “we are manufacturing batteries” is the full answer. We are manufacturing factories that make batteries because of the amount of demand that is coming. All the environmental goals institutions planned won’t be achieved if there is not enough supply. We can’t tell people: “Sorry, this electrification thing is going to take longer than we thought.” This transition is so huge, and the only way we can meet the challenge and the demand, the only way not to slow it down is to be able to scale faster than any industry has ever been able to do before.
The demand for batteries in Europe in the next 10 years will grow tremendously. Northvolt is a company that is manufacturing batteries by manufacturing facilities. We are a factory for factories. We need to automate things inside the factory, but also the process of building new factories and find ways of making the factories we have less costly.
We’d have a big scale problem that could constrain our demand in the market if we can’t figure out how we apply new principles to designing, building and operating factories that look more like what we see in other high-tech areas, where the level of automation is much higher, especially in software.
At the core, I think this is a data problem. It’s an information problem. Once you have the information, the physical part becomes so much easier to automate.
Why did you decide to leave Tesla for Northvolt?
Several reasons. The team: I was incredibly impressed by the team. The role: it was exciting for me having the company share this vision of automation as an information problem, as something we need to invest and focus on. The project: it’s such an audacious, huge, bold project. And it didn’t hurt that the job was in Europe. I really like Stockholm
What do you and your team do at Northvolt?
Automation at Northvolt combines IT, software engineering, controls engineering and robotics and material flow.
As a company, we want to collect a tremendous amount of data, so back to this idea that automation is an information problem. If our goal is to improve the process as quickly as possible and as efficiently as possible, we need to give the system enough information to know what’s the right decision to take in the process.
We collect all of this tremendous amount of data from how a product is made, but usually we lack information on how it ends up. We need to be able to talk to the battery after its gone and track how it’s doing out in the real world.
We bring all of this data back together, and it tells us that when we make a battery in a certain way, this is what will happen later. We can tweak the battery, and make it better. Or we can tell the customer’s system that if it changes the way it charges or discharges the battery, it will have a lot better outcome. Or we can update the BMS software that manages the battery firmware so that the management algorithm is more efficient for that specific battery.
We are also trying to build open source things as much as we can, especially on the software side. We don’t want a culture of closed information. If you get more people to collaborate, more people can see and run into the same problems. By sharing the work, something that might take 10 years to build can take instead three or four years or less. It can enable entire industries to move this fast too. When we are solving problems that are common in all sectors, we shouldn’t be the only ones having the benefit of that. We’d love if other companies want to partner with us to help us do it. If we are going to solve the same problem as you are, why don’t we solve it together?
Automation, software engineering… what do you do in your spare time?
I’ve got two adorable kids and the most patient wife in the world, so almost all of my time goes towards hanging out with them, which is the way it should be and couldn’t be spent in a better way.
I’m a nerd, so I like to play video games. I have a six-year-old son and a three-year-old daughter, and I’m super psyched that they both like video games too. We play a lot together.
I used to rock climb a lot, but I haven’t done that very recently. Probably it will be something I’ll start doing again with my son, so hopefully we can get more into that together. I also like to cook a lot on the weekends.
I definitely wouldn’t call myself a chef. I can reliably burn food, especially pizza!