Military Embedded Systems

AI technology and USSOCOM


May 05, 2024

John McHale

Editorial Director

Military Embedded Systems

Pictured: Shield AI's MQ-35 VBAT

TAMPA, Florida. Shield AI engineers are working to change the way the U.S. military and U.S. Special Operations Command (SOCOM) leverage artificial intelligence (AI) capability. In an interview I did with Brandon Tseng, co-founder and president of Shield AI, for the SOF Week Show Daily, he and I discussed why USSOCOM is the best customer to deal with when introducing new technology, the tech Shield AI is showcasing at SOF Week in Tampa, and how AI pilots will enable small groups of warfighters to command the same combat power as a carrier strike group. Brandon also shared experiences from his years as a U.S. Navy SEAL. Edited excerpts follow.

SOF Week Show Daily: What is your role at Shield AI and your experience in the defense industry?

Tseng: My name is Brandon Tseng. I'm the president and co-founder at Shield AI [San Diego]. I spent the first seven years of my career in the Navy. Starting off as a surface warfare officer, an engineering officer on board a ship, I was a mechanical engineer. Out of the U.S. Naval Academy, I immediately went to the USS Pearl Harbor, LSD 252, deployed to the Persian Gulf. Then I got my Surface Warfare Officer pin and later transferred to the Naval Special Warfare community. I subsequently became a Navy SEAL, and I deployed twice to Afghanistan, and once to the Pacific – Afghanistan with SEAL Team 7 and the Pacific was SEAL Team 5 as a procurement commander.

SOF Week Show Daily: Why is the Special Operations community important to Shield AI?

TSENG: Because they're important for all the armed services. SOCOM [United States Special Operations Command] has always been at the tip of the spear. They've been early adopters of technology. They are focused on solving problems in the fastest, most expedient, most efficient way possible.

So, when you are introducing new technology, new products, it's hard to actually find a better customer than SOCOM. There is a tolerance for risk. There's a tolerance for a level of failure. There is a desire to move quickly to get capability out to the warfighter and get that user feedback from a product-development cycle so we can make the product better and better and better.

That's why SOCOM is a fantastic customer for Shield AI. Because it's hard for me to point to other services that are willing to go through that product-iteration cycle closely with a company. We've had great success doing it with our quadcopter product-development cycle working hand in hand with SOCOM operators to make the best product possible, product requirements be damned.

We're also working with the SOCOM customer on our MQ-35 VBAT [vertical takeoff and landing uncrewed aerial system], product, and Hivemind [autonomous artificial intelligence pilot] as well. [See image at top of article]

SOF Week Show Daily: What tech will you be showcasing this week at the event?

TSENG:  At SOF Week we will be showcasing the MQ-35 VBAT. We will be showcasing ViDAR [EO/IR AI system] which is a recent acquisition of ours from a company called Sentient Vision Systems. And we will be showcasing Hivemind or VBAT teaming with Hivemind being our AI pilot, which makes teams of VBATs possible.

SOF Week Show Daily: What is the biggest difference when working with SOCOM as opposed to other parts of the military?  Is it faster procurement? More direct input from the operator/end user? Something else?

TSENG: It's that closeness in terms of the product iteration in the development cycle. That’s what comes to mind, access to the operators: A true team effort between industry and government to get capability out to the warfighter. That's the amazing thing about SOCOM. Some places only talk about it; with SOCOM we are working directly with the operators to best understand their needs, to best build the technology and capabilities that they need. That's going above and beyond any formal program requirements. Just because you meet program requirements doesn't mean you have a great product. You have a great product when operators say you have a great product. You only get that when you're working closely with the customer base and that's where SOCOM really shines.

SOF Week Show Daily:  I’ve had operators over the years tell me they often don’t see the cutting-edge tech we see at shows like SOF Week in the field. As the military adapts commercial tech more quickly, is that changing?

TSENG: When I think about my experience with technology in the field, we had the things that were issued to [us]. Probably the most advanced piece of technology that we had was a SATCOM radio. We also had great optics on our weapons. We had fantastic radios to communicate with aircraft. And we had good intelligence products that were generated from satellite imagery and a lot of intelligence gathering.

When I think back, I was never bothered by the tech I didn’t have. As a warfighter, you're issued your equipment, you're trained to fight with that equipment, and be quite effective with it. I remember the first time in 2014, when they handed me a cell phone and they said, Hey, this is called ATAK [an app called the Android Team Awareness Kit (ATAK)]. We want you guys to do Blue Force/Red Force tracking on it. My gut reaction was like, well, I have two years to be an outstanding platoon commander. I am very good with Blue Force/Red Force tracking with my laminated map and my Sharpie erasable marker. That's how I do it and I can do it incredibly rapidly and that's what worked for me during combat deployments to Afghanistan. I'm not going to take the time to learn how to use ATAK because I fundamentally had two years to be a good platoon commander and this is going to inhibit my ability to be a good platoon commander.

That was not necessarily the typical thinking, but not atypical in terms of warfighters on the ground. Technology needs to be introduced and fielded in a way that makes sense.

Years later, I remember going back to the SEAL teams when [we did] testing of our quadcopter and I asked them, “Hey, do people use ATAK?” And [they were] like, “Oh, my God, everybody uses it and it’s bonkers if you’re not.”

But I also think this generation of younger SEALs have been brought up in the technology age, they more quickly adapt to it than my generation. I think of my three-year-old daughter and wonder how quick she’s going to be at adapting and using new technology.

There is also an adoption curve. It takes time to integrate and train operators on the new tech. There also needs to be realistic expectations. The tech is much improved and state of the art, but not what you will see in a Captain America movie. That said, tech is coming more quickly now.

SOF Week Show Daily: Where do you see AI impacting Special Operations Forces in the military five or 10 years from now?

TSENG: AI pilots, a self-driving technology as applied to drones. These can be UAVs [uncrewed aerial vehicles], UGVs [uncrewed ground vehicles], you name it. It doesn't matter which domain they're operating in. This is something I speak a lot with international customers about, but there is no reason why a SEAL platoon or a Ranger troop can't wield the same combat power as a carrier strike group. No reason why with AI that those 16 people can’t have the same combat power as a 5,000-person carrier strike group consisting of 200 aircraft, short- and long-range missiles, destroyers, etc.

That is something that I think is so important for people to wrap their heads around in terms of the impacts of AI pilots. It’s coming, and it’s coming quickly where you will have a single person capable of commanding 300,000 aircraft loitering munitions, effectively with the optimal effectiveness on the battlefield because of the AI that is piloting, commanding, and maneuvering these assets.

It's fundamentally a massive paradigm shift. SOCOM has the smallest budget out of the services. There is no reason why they can't have the most impact if they employ those resources in a sophisticated way as it relates to AI and autonomy.

But that's the same thing I tell international [users]: If you're the British with a $50 billion defense budget or the Germans, there's no reason – if you resource allocate effectively around AI pilots, around autonomy – you can’t have the same impact as a military that has a $900 billion budget. That's the power of AI and autonomy. It’s a force multiplier on steroids.

SOF Week Show Daily: Can you elaborate more on how that small group can have the same power as a carrier group?

TSENG: There's a couple of aspects of it: One is when commanding these aircraft, there's no reason why an E6 special operator can't command the number of assets that would otherwise be commanded by a four-star general – and arguably with perfect effectiveness, and that's the power of AI.

Then it becomes a matter of production. Now you have these hyper-intelligent aircraft, these hyper-intelligent drones, loitering munitions, and you're not limited by the force size anymore from a personnel standpoint. You're simply limited by a production capacity.

That's what I mean by they can have the same combat power as a 5,000-person carrier strike group. You do that in the right way [with AI] and one person can command all of that mass, all of that firepower, in the same way that a fleet admiral does with the carrier strike group.



Featured Companies

Shield AI

600 W Broadway
San Diego, California 92101