Vaibhav Kumar Georgia Tech

Fast 5: How A Student-Created Startup Is Revolutionizing Logbook Research

Vaibhav Kumar, an aerospace engineering graduate student at Georgia Tech and one of Aviation Week’s 2019 20 Twenties winners, talks about founding startup AirLogs—which seeks to save mechanics time by digitally analyzing maintenance records.

What got you interested in aviation—and in particular, aviation maintenance? 

In 2017 I was flying and I had a partial engine failure at 200 feet above the ground, which is kind of like the dead man zone, so that instant propelled me into aircraft maintenance. When we got back I realized what the problem was—it was a spark plug failure in one of the cylinders—so that pushed me into understanding why that happened and what I could do to make sure that never happens again, to anyone.

I decided I was going to build an advanced engine data monitor that will measure a spark’s health over time and kind of predict when the spark would fail so that this specific problem wouldn’t happen. With that idea in mind, the next semester I took courses in internal combustion. I spent that winter break working at a repair station in Atlanta trying to see how mechanics work and think, and in that process I ended up logging more than 100 hours of supervised maintenance. I spent the next semester studying internal combustion and working on a new digital engine data monitor that we can use for our fleet of airplanes.

You serve as a maintenance crew-chief for the Yellow Jacket Flying Club at Georgia Tech. What is that like?

Because I was so involved in maintenance, it was a natural transition for me to stand up and lead as the crew chief for Georgia Tech’s fleet of airplanes. I’m the crew chief for three of the four airplanes. Georgia Tech has four Cessna 172s, which fly a total of 4,000 hours per year, so it's a really busy flying club. Because we are a nonprofit, we try to keep costs down and we usually have our own mechanics and our own crew chiefs who are responsible for maintaining their airworthiness. As a crew chief you’re responsible for doing weekly inspections of the airplanes, keeping an eye on oil consumption, changing oil, changing oil filters, changing spark plugs, cleaning spark plugs, lubricating control surfaces—all the things that federal aviation regulations allow you to do [without being a licensed mechanic].

Your research into building an engine data monitor led you to co-found a startup called AirLogs. What has that journey been like and what is the goal of AirLogs?

Georgia Tech has an incredible startup program called CREATE-X and they told us about this process called customer discovery, where you go talk to customers to understand what their problems are and then build the product, instead of building the product and then trying to sell it—because you want to be sure there’s a hard market for it. So, following that advice we went out and talked to 160 people in aviation, including mechanics, repair station owners, aviation inspectors from the FAA—all kinds of people. Over the course of that process, we went through a couple of pivots. We moved from building an engine data monitor to maintenance analysis, to predictive maintenance, to analyzing aircraft maintenance records. It’s analyzing the aircraft maintenance records that really got people interested enough to call us back and ask us if our product was ready, because they wanted to buy it. That was the indication for us that this was the way we wanted to go: we had people offering money so we knew this was valuable.

What AirLogs does is basically digitize, analyze and research aircraft maintenance records. Airplanes are legally required to have aircraft maintenance records. We scan them, and once they’re digitized we use machine learning to analyze the typewritten information and back out [an aircraft’s relevant] airworthiness directives, service bulletins and project documents, what [maintenance tasks] were done, and the bill of materials for the aircraft, so we know what needs to be done in the future. We essentially determine everything that has been done on the aircraft and that needs to be done on the aircraft, and this has two major value propositions.

Right now, mechanics spend 30% of their time doing this analysis, and this analysis is literally looking at books and checking different sources—like the FAA website, for instance—to see what needs to be done for this specific kind of engine of this configuration, and what has already been done. They’re not really turning wrenches, they’re just looking at stuff on a computer and backing out things on a logbook. Secondly, because we know the entire history of an aircraft, we know the value of the asset that you own, so that’s another market that we’re trying to get into.  

How close are you to bringing AirLogs to the market?

We’re still developing it. We lost one of our main leads because he got recruited by Google, and we’re pivoting a little bit away from just doing general aviation maintenance because we were trying to raise money for our next round of funding. With CREATE-X we got our first round of funding and now we’re trying to raise money for our next round, and investors aren’t extremely happy with the market value of general aviation. We are in the process of a pivot, so our product will be quite different from what it is right now, but we do have a prototype.

We had plans to do our pilot study the end of 2018, so we had two repair stations that have signed up for a pilot study and that was our main plan, but since we recently lost our co-founder, that plan has been pushed back and we’re actively recruiting to get another person on board to launch our product in the next two to three months. We’re planning on doing it by the end of June, but I have a feeling that’s going to get pushed back because of all the directions we’re getting stretched into. With a pilot study, we’re isolated to a certain kind of airplane for a certain repair station, and once we work with them we’ll be able to figure out if our business model is valid. If it is, we’ll be able to take that information to potential investors that we’re talking to and see if they’re interested in funding us. That would help us scale to different repair stations and go from there.

What’s next for you? What are your future goals within the industry?

While we were struggling with funding issues, I ended up picking up a role at Delta. I’m building Delta’s next engine planning system, which is how Delta will be planning how it repair engines—how it forecasts removals and staggers them across their fleet.

In the next few years, I want to push AirLogs and change how we track maintenance in general aviation. I think there’s lots of value there. Ten or 15 years down the line, once we get this new breed of eVTOLs entering the ecosystem, we’ll have to develop maintenance standards for them. We’ll have to build something that will work to scale, because right now general aviation is pretty small, but with the main concepts they have, general aviation will get pretty big. So we’ll have to come up with maintenance procedures that allow for that scaling. Once this change happens, our turnaround times for repairs have to be as good as turnaround times we have for cars, and we are pretty far from that.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.