Tuesday, March 12, 2019
Interview with Bryan Biniak, Connected Travel
Two of the hottest sectors in the market today for investments and startups are in the area of machine learning/big data and advancements in automotive technology. At the intersection of those two is Los Angeles-based Connected Travel (www.connectedtravel.com), the startup which is behind Honda's Drive Drive platform. We caught up with founder and CEO Bryan Biniak to hear more about the company and what Connected Travel is doing.
Tell me about Connected Travel?
Bryan Biniak: Connected Travel was founded in 2016. We have built an application services platform for vehicles, and our first customer was actually USAA. We built a behavior-based insurance platform for them, which is still operating today, which uses AI and machine learning technology to understand people' day to day driving behaviors and routines. The system can do things like learn where you have driven, all the places you go to- and from- every day, the routes you take, and long you stayed there, and other behaviors related to those things. It can figure out when you go to work, when you're early or late for work, and the routes you take. In that context, we can figure out how weather and traffic impacts your driving routines and whether you'll be on time or late, when you are speeding, when you're using your phone in the car, and more. We built a whole, loyalty rewards and points system to incentivize changes, and also helps with behavior management and sustained change. For example, you can earn points for reducing your speeding, and those points can be redeemed at bricks and mortar and online retailers, and also other reward programs. It's been fascinating to learn about people's driving behaviors, since that's what 90 percent of the people are using to get around. Less than 10 percent of people are taking public transportation, and only about 20 percent of commerce is online. Through this insurance journey, and using behavior-based insurance, we opened up this world around commerce. That's where we started to play with Honda, our next big customer, where we have built out these infotainment and commerce systems in a vehicle, with an ad platform on top of what we started as an incentive for safe driving. We're applying that same platform and are now using it for information, commerce, and advertising in vehicles.
How did you end up working in the whole area of automotive AI?
Bryan Biniak: I was back at Nokia running the products ecosystem with the app store. We had an advanced application development and innovation team. We had a new tablet we had developed back in 2007, and we were asked to come up with concepts to drive cellular usage. Most people were using tablets with Wi-FI only, and they wanted to target families and be able to sell tablets and smartphones at the same time. So, the only place we figured out where Wi-Fi was not widely available was in the car. Nokia owned Heremaps at the time, and we had a couple of other people working on contextual content and imaging, and we were able to adapt to augmented reality some ideas we had had with Dreamworks Animation, and their movie, How To Train Your Dragon 2. We took all of the data we had, including telematics from the car, from the phone, mapping and POI data from some companies like Factual, 4Square, and other social data, and even data from Flightaware and weather data from the Weather Channel, and fused them all together to build a unity rendering engine. As you were driving, you'd be able to see that 3D animated world as you are driving, look up in the air and see a plane flying, and see whether that was a flight from Southwest Airlines, United, or American Airlines, and we'd render dragons. We'd do things like turn the 405 into a dirt road and muddy road. We gave these to clients in 190 countries around the world, so they could experience this imaginary world, and we built the application for kids because the parents should be driving.
The idea was that it would gives kids and parents a way to interact, an application so people could play with their kids while driving at the same time. That was just the beginning of it, and we used that to build a next-generation flight tracker, then we connected that with the largest network of commercial buoys and plugged that into the platform, letting Royal Caribbean render the undersea world in real time. So, we had land, air, and air and sea, indoor and outdoor. Along the way during that process, Nokia got acquired by Microsoft, and I ended up becoming a General Manager at Microsoft. At some point, something more compelling and interesting came up, and I left and went back to Nokia as an EIR in their venture capital fund, focused on mixed and virtual reality and IoT. We got involved in a project with Visa, building a next-generation concierge service using real time transaction data, and then this opportunity with USAA popped up, which became the catalyst for starting the company. It's been exciting, being able to start a company with revenue, as opposed to venture capital money, and to build the company the old-fashioned way, with revenues. We've done very well, and are funded 100 percent with revenues.
What has the consumer response been to all of that tracking of their data?
Bryan Biniak: They're a clear value exchange, in that they get points and rewards for that. A lot of the time people understand that it's innocuous. For Honda, in particular, they understand we're using gauge information, things like your odometer, HMI data and other sensors, and we tell the customer up front that we're using it for this purpose, to specifically create value for you, and by the way, we're going to give you something in return for that data. The more you use it, the more points and rewards you get for it. None of the companies we're working with are taking that consumer data and selling it. They view it as their customer's data, which is very simple and straightforward. The idea is to provide greater value and enhance their everyday travel in their care, help them become safer drivers, and in exchange for that, we give them something.
What are some of the challenges of big data and machine learning today you have encountered?
Bryan Biniak: We've built everything from the ground up on our side. I think it depends on the source of where you're getting your data. If you're getting it from the vehicle, the challenge is that not every vehicle has the ability to provide the same data. Not every vehicle has the ability to provide that information, in a common format, at the same level of frequency. That makes it very hard. There are 218 million licensed drivers in the u.S., and all are driving cars at different moments—but there's only a very small set of those that are able to provide actionable data. We're very focused on real time, actionable things we can do to create value for a customer while driving. We're not interested in much of what happens in the past, mostly in what's happening right now and in the future. So, when you talk about using this data, AI, and machine learning, the question is how you create that framework at scale. That's the hard part.
You have to fuse together lots of data sources to get there. Some are very clearly accurate, and others are more simulated, because you're stitching together information from a phone, or other IoT devices, to augment what the vehicle can or cannot provide. It's been interesting, we've been working with lots of radio broadcasters, and once you start understanding what's happening from an insurance standpoint, such as people's behavior and routes, that's mean we could start working with radio broadcasters, and you can start to not just know what song they are hearing, but what advertising. Since you know a person's route, you also know what billboards they are driving past. If you think about it, advertising is trying to get you to do two things. Either, they want you to go online and buy something, or they want you to get in your car and drive out and buy something. The vehicle really has been, for the most part, the most unintelligent shopping cart that exists. However, if you look at what happened online with shopping carts, that's what hasn't yet happened here, but what we've been able to start to do.
With Honda, one of the things we built with them with Honda Dream Drive, and initially with CarPlay and Android Auto using Siri and Google Voice, is a retail interface to the head unit. So, when you're driving, you can find parking on the street or at a garage, and there's a whole payment system built into the vehicle. We also worked with a bunch of big gas station businesses, including Chevron and 76, so you can find gas by price or by location or brand, pull up the pump, tell the car your pump number, and it will take care of that transaction for you, so you don't have to use a credit card. We added Yelp, so the car can find you a restaurant and book a reservation. You can order ahead, and we're now working with Atomticket to allow you to order movie tickets, with Coffee Bean and Tea Leaf to order you coffee, and lots of other things you have wanted to do while driving but where trying to pick up a phone and navigate all of these menus might not be safe. Going full circle, is how do we build in safe driving behavior, and create an interface that allows you, with people spending an increasingly large amount of time in the car, but be able to use that time productively. We want you to be able to do a lot of things you typically wouldn't be able to do if you're driving safely, or where you might normally have to wait until another time to do it. We want to make your drive time much more productive.
What is the most challenging thing you having to tackle now in your business?
Bryan Biniak: The most challenging thing we're tackling, is deciding, as we grow and add more customers, how we grow the business at the same time. We've really gone out of our way to see how far we can take this growth, purely on revenue, and at the same time, we're seeing a big opportunity to go faster. There are 56 auto manufacturers out there with CarPlay and Android Auto enabled, but most only have audio and mapping. We have the ability to provide a full services and commerce interface. You have to be in a lot of places at one to support all those customers. I think our biggest challenge, is do we keep growing customer-by-customer, or do we accelerate that growth, raise some venture capital, and grow the business ahead of all those deals. That's probably our biggest challenge.
What is the biggest lesson you've learned so far with this startup?
Bryan Biniak: It's patience. We started this in late 2016, and we've just kept our heads down for the last couple of years, building up the business and making sure we're successful with our customers. We didn't really talk about the company until Honda launched its platform at CES, and provided us with attribution. Since then, it's been great. It's really let our work speak for us, versus us putting out press releases and all the pomp and circumstance, which is historically how lots of people have tried to draw attention to themselves. They do that to get money and use that to build something, whereas we did the opposite, build something, get customers first, build our service, launch it, and grow it, and let the product speak for itself. That's very different than what I've done in the past, and it feels a lot more satisfying, that's for sure.
Thanks, and good luck!