Lumo, a Boston-based travel technology startup, is telling airlines it can peer into the future and know which flights are likely to be delayed hours — or even days — in advance.

Three years into its life, the company — which helps airlines and corporate travel managers preventatively rejigger routes and itineraries — said Monday it had received a $2.3 million seed round.

Frankly, the dollar amount is unimpressive. But the companies backing the startup are notable.

Leading the round is JetBlue Technology Ventures, the San Carlos, California-based investment arm of the U.S. airline JetBlue Airways.

Another investor is Founders Factory, a London incubator and accelerator whose travel program is backed by UK airline EasyJet. Lumo has been mentored in the program this winter, which is where it gained insights into the European travel market and regulations and received help with marketing, such as by rebranding from FlightSayer to Lumo.

Plug and Play, a Silicon Valley-based accelerator that has worked with many travel startups, is also an investor. This spring, Lumo will receive mentorship there.

The latest investment round also includes a stake from an unnamed strategic investor, which is likely a major airline or a travel management company.

Already the flight delay prediction startup has signed up travel management companies Adelman, Flight Centre, Carlson Wagonlit, and Acendas as pilot customers.

These companies want help rebooking flexible travelers when the chance of flight disruption is greater than 75 percent based on various factors including historical performance, seasonal trends, the time of day, day-of-the-week, and airline track records.

In one test, a travel management company was overseeing travel for an event with about 1,000 guests when a winter storm hit. Lumo’s web-based dashboard showed the likelihood of the event’s speakers and other very-important-people having their flights canceled. That information enabled the travel managers to put into action back-up plans.

Consumers may be interested in Lumo’s predictions, which are free online. Type in your travel date, flight number, and airline for a trip within the next few weeks and the company will offer an easy-to-understand prediction of the chance of different lengths of delay.

Flight Delay Crisis

In January 2018, for example, 18 reporting U.S. airlines had one out of five their flights arrive outside of 15 minutes of their schedules, the U.S. Bureau of Transportation Statistics said.

That on-time rate was roughly as poor as the same month in the past few years. It also represents a lot of irritated flyers — especially as a result of so-called Nor’Easters, or winter storms in the northeastern U.S., that prompted many delays and cancellations this winter.

With Lumo’s analysis, travel management companies may flip that story without needing to make tradeoffs on their bottom lines.

Travel managers can rebook passengers, earning customer loyalty that justifies their overall fees. They can also minimize drama at the gates. That drives down labor costs because agents spend less time rebooking passengers at the last minute — when alternatives are fewer.

Rocket Science

Another impressive aspect of this otherwise humble startup is that it has received $1.75 million in grants from the U.S. National Aeronautics and Space Administration (NASA) to model airspace efficiency between 2015 and next month.

That work lay the groundwork for Lumo’s broader business, which uses machine learning and reams of data to make forecasts.

The space agency was impressed by the street cred of the company’s founders. CEO Bala Chandran earned a Ph.D. at the University of California, Berkeley. The full-time chief technology officer is Diana Pfeil, who earned a doctorate from MIT.

Forecasting a Startup’s Future

For carriers, Lumo holds promise in that it might help them streamline their operations for potential cost savings, with fewer gate agents needed on the front line of customer support.

On the enterprise side, it risks being outgunned by the resources of established players. Its lack of automation and of integration with travel management companies tools are also drawbacks preventing rapid scaling up.

SITA FlightPredictor attempts to offer a similar service for airlines. SITA, a Geneva-based organization owned by the air transport industry and formally called Société Internationale de Télécommunications Aéronautiques, also uses sophisticated math and software to predict flight disruptions.

Since 2016 Amadeus has offered airlines a so-called schedule recovery system with Qantas as the first airline to try it.

On the consumer side, Google has, since November 2017, been rolling out flight disruption alerts via its flight search tool.

It’s unclear how Lumo, which only has a dozen employees, can compete.

The company also hasn’t put its predictions to the test against a third-party analysis. And while being an improvement over guesswork can help, the company still might miss an enormous number of delays. Everyone should beware of hype potential.

Another startup, Cambridge, Massachusetts-based Freebird, provides technology to cope with a disruption once it has been identified. (It offers some predictions.) Lumo doesn’t offer a fulfillment tool like that, in comparison.

The startup’s introduction to airline executives through accelerator programs also may not translate into deals. Neither JetBlue nor EasyJet, for example, operate network airlines with the types of complexity that are truly prone to disruptions. Neither carrier has signed a commercial agreement or doing a test pilot with Lumo.

Chandran countered that his company is doing a pilot this spring with a major international airline and that — if the test succeeds — the carrier will become a reference customer.

Hubris can harm many prediction-based startups. All it takes is one Icelandic volcano to spew up ash clouds and ground air traffic control to a halt.

Chandran admitted that his company’s predictions are much more valuable for routine disruptions that, while not ending up on CNN, can cause business travelers to miss flights and meetings. “Here in Boston, you can look out the window, and it seems sunny, but the fog at the airport may lead to reduced visibility and cause delays,” Chandran said. “We can anticipate that.”

That may be so, but the company risks overstretch by trying to appeal to both travel management companies and enterprise customers along with the development of consumer mobile apps. A consumer app for iPhone and one debuting shortly for Android devices offers alternative flight suggestions when delays happen.

Yet if Lumo can pull off its predictive tech accurately and cost-effectively, the president of JetBlue Technology Ventures, Bonny Simi, said that would mark a significant change. To learn more about the investment, hear Simi talk at Skift Tech Forum on June 12.

In the meantime, we all have to wonder how long it will be until artificial intelligence can predict which startups in this sector will succeed.

Photo Credit: Flights canceled at London's Heathrow during winter storms in December 2010. Lumo is a company that has received investment to build better predictive tools to avoid such flight delays. Sean O'Neill / Skift