Defense Tech Spotlight: Tern AI
Episode 9
I had the privilege of speaking with Shaun Moore, Co-Founder and CEO of Tern AI. Here are the highlights of our conversation:
What is Tern AI?
Tern AI is developing an independently derived positioning system (IDPS™). It is a software solution catered to the land navigation market and decouples the US economy's reliance on GPS. One way in which we're reducing the GPS risk is by providing position information that is independent of physical infrastructure such as satellites. We can determine position, entirely self-contained within the individual unit or vehicle, without the use of signal.
Could you walk me through your background and how this startup came into existence?
I spent about 11 years in the facial recognition space, so my professional career has largely been focused on developing and commercializing AI. The Tern AI position recognition software is capable of recognizing where you are on the map. Many of the lessons learned and knowledge that I spent 11 years building in AI are applicable to what we're doing at Tern.
I sold Trueface in May of 2021 to Pangiam, and then I spent two years as an executive and board member there before joining Tern AI as Co-Founder and CEO in 2023. I am extremely enthusiastic about the magnitude of the opportunity and our ability to provide a much-needed solution to such a widespread and important national security concern.
Where did you get the idea for Tern AI?
My Co-founders have many years of professional navigation experience. Looking at their first-hand experience and perspective, we started talking about the advancements in AI and the possibility for it to have a significant impact on our critical infrastructure, specifically GPS. The research we were pursuing was to understand how organizations were innovating within the positioning and navigation industry.
We found most organizations are attempting to improve upon what already exists, however our foundational principle was to not use any external signal, which is a radically different approach on how to solve this problem.
Can you talk a little bit more about the technology behind Tern AI?
Tern AI’s proprietary navigational brain utilizes digital maps and smartphone/vehicle data to recognize and derive the user/vehicle’s precise position.
So, there’s no reliance on GPS, if the satellites were to go down for whatever reason, Tern AI would still work and give directions?
Yes, Tern AI will still be able to provide position data for directions. We're providing the positioning input for turn-by-turn navigation without any reliance at all upon GPS or external signal.
For example, Tern AI wouldn't replace Google Maps, you would partner with them to only provide positioning information?
That's right, Tern currently only supplies position data and can integrate with potential providers such as Google Maps in place of the GPS input signal. It's providing significant resilience to the economy if something does happen but also providing increased efficiency, no more lost signal, and no more missed turns.
In society today, our economy’s reliance upon turn-by-turn navigation is ever growing. If New York City’s GPS is jammed imagine the mayhem it would cause to the economy and to everyone's livelihood. Our goal would be to partner with the Map providers so that everyone has access to continuous and accurate information regardless of what is happening in space or with our adversaries.
Does your technology only work on roadways, where it can recognize the road network? Or do you also imagine this working for vehicles traveling off road?
Great question. Our near-term focus is the mapped road network, but that's absolutely a step down the line. We've been thinking a lot about off road, and we think we know how to solve it and will focus resources on it when the time is right.
Is this where your background in facial recognition and pattern matching comes in handy? At a simplistic level, is the road network the face and you're just matching where the vehicle is on the face (road network) to get location?
We call it our position recognition technology. The roadways have very unique features as you traverse them, similar to how a face has unique features. When you apply the Tern AI navigational brain to make a prediction in near real time, the result is very precise and continuous positioning that is not subject to interference.
Do you have to have a suite of sensors on the vehicle to be able to see where it is on the road network?
We are able to use the existing infrastructure in phones and vehicles, if necessary, we can retrofit older vehicles.
Where/how does AI play a role in all of this?
We apply our proprietary AI to predict and confirm position information with all of the available data we have access to. To put it simply, its purpose is position recognition.
Who are some of your competitors in this space and what you guys are doing to differentiate yourselves?
Our approach is not to improve what's out there. It's to truly provide a breakthrough in positioning technology and remove the bullseye from the back of GPS. As AI, and even quantum computing continue to gain momentum, what we believe we end up seeing is a matrix of PNT with air, sea and land. So you have position, navigation and timing which will each break into its own niche and then you have land, air and sea, which may or may not use the same solutions.
Right now it's all coupled into one risk, one significant risk for the United States economy. I view this as much more of a collaborative effort by the industry to solve the problem, and I think it's important that collaboration exists because this is a national security issue, and it needs to be solved. From the competition side, going back to our founding principle, it is no signal. That's very different from most. Our most important differentiators are that we provide a software solution that can be deployed overnight, without needing access to any external signal or significant infrastructure investments.
I think it's important that we're seeing competition pop up because we, as a country, have to solve this problem.
You mentioned the sea, air and land components. Are you solely focused on land? Do you think Tern AI will eventually branch into those other components?
We're seeing companies pop up, primarily in the GPS denied UAS space. I think that we will explore this area in the future, but at this point we are laser focused on executing on land navigation.
I saw you recently announced a partnership with IBM, congratulations! Who are some of the customers that you see for this application?
Yeah, it's great to have a name like IBM supporting this initiative. They're helping us with our data pipeline. We are seeing a lot of interest from logistics, emergency services, last mile autonomous delivery, rideshare, and so we are pursuing those very actively.
Are these companies choosing to work with you because they are concerned about the denied environment and they want a navigational backup that is not GPS reliant, or are there other reasons?
That's a component of it, operational efficiency is also of significant interest. When you consider logistics companies or even emergency services, it is imperative that you have continuity and precision with your position information. Right now, the only thing that exists is GPS, but it's proven to not always be reliable or precise. This is the first time that an easy to deploy alternative to GPS exists.
You talked about the traction you’ve had on the civilian side, have you had much on the military application of your technology?
We are actively pursuing both the commercial and Federal markets. We were recently awarded a contract by the Department of Transportation for a Complementary PNT demonstration which is a big step in the right direction. In addition to the DOT, we are pursuing Alt PNT for the Space Force and Autonomous Conveys for the Army. We believe that our solution would be a great way to ensure that autonomous convoys can operate in denied environments without exposing operators to risk.
What's been the most challenging aspect of building Tern AI so far?
I'd say the most challenging aspect has been entering this industry, truly from the outside, and taking a very different approach than what people are used to. GPS has been around for 50+ years. The resistance in driving innovation in the industry has been challenging, but we're overcoming it by demonstrating our approach works.
Outside of that it's really been making people aware that this issue is truly a large national, and global, problem because of the risk that comes along with it. And I don't think that the public has really thought much about it, but it truly is a concern.
Do you envision a symbiotic relationship with GPS or a complete replacement of GPS?
I think in the future we are in a multi-stage PNT effort, where you have multiple solutions that can all work together. Our goal is to be on every single vehicle and to be the de facto position technology. I don't see the US government turning satellites off and taking them out of the sky, that's not realistic in my view. My view would be that eventually GPS supports the other technologies that are out there that have been making significant strides in innovation.
Thanks for reading!