When GPS fails: this small cap is fixing a $54B drone problem

Key Points:

• Sparc AI (OTC:SPAIF / CSE:SPAI) is tackling a growing global problem: drones and aircraft failing when GPS is jammed or spoofed, a problem already driving massive government losses globally.
• Its Overwatch platform is a software solution to the hardware problem, using AI and existing sensors making it highly scalable across fleets.
• Early traction includes trials in the UAE and India plus expansion into the U.S. market to tap into the world’s largest defence spending base.
• Sparc AI expects three revenue streams: annual fees for Overwatch, annual fees per drone, and API fees tied to its AI capabilities. Overwatch is expected to have a 95% gross profit margin.
• At a ~CA$66M market cap targeting a ~$54B market, even modest adoption could drive meaningful upside. 

The Problem: drones are expensive and too easy to shoot down  

Last week, the world received news that the US’s total drone losses from the US-Iran War is now approaching an astounding $1 billion.  

Its latest loss of the MQ-4C, a surveillance drone worth $240 million on its own, was an especially big hit alongside the 24 other MQ-9 drones (worth $30m each). All of them have either crashed, been shot down, or perhaps… just lost somewhere. 

Defense experts are saying it’s likely Iran is able to disrupt drones through GPS interference. And not just drones - civil aircrafts, fighter jets and all other air traffic that rely on GPS get disrupted or spoofed. Since 2024, there’s been over 430,000 jamming and spoofing incidents for just aircrafts. It’s been a massive (and expensive!) problem for leaders all over the world. 

So as a tech enthusiast, I naturally went to look for the company that could give the best possible solution to this problem because I can imagine a strong demand for it.   

I looked into the big names - Collins, Honeywell, Safran Electronics, Thales - to name a few. But interestingly, I found the solution to be from a rather unknown, sub-$100m market cap company.  

Introducing Sparc AI 

Sparc AI (OTC:SPAIF / CSE:SPAI) is a defense tech company trying to solve this exact problem - what happens when GPS stops working or is tampered with. 

For those unfamiliar, most drones rely heavily on GPS to understand where they are. When that signal is disrupted or unavailable, they tend to have internal sensors to fall back on but these aren’t very reliable on their own.  

Solving this problem traditionally means adding more hardware - things like radar, lidar, laser or other high-end navigation systems. The issue is, these solutions are expensive. They can cost hundreds of thousands of dollars per drone and may need extra hardware that increases the risk of damages. 

Sparc AI takes a completely different approach. Instead of adding hardware, it uses software. 

Sparc AI’s Overwatch solves the problem  

Overwatch, Sparc AI’s brainchild, uses Machine Learning to optimize existing sensors inside a drone to provide precise position without GPS. Here are the key points on how it works. 

Overwatch is a software-first solution to a hardware problem. Instead of relying on expensive sensors like lidar or radar, Overwatch leverages a drone’s existing hardware and makes cheap sensors behave like high-end systems.   

At its core, Overwatch fixes bad data. When GPS drops out, drones quickly drift and lose accuracy. Overwatch uses machine learning to analyse data from the drone’s sensors to correct errors in real time, keeping the system accurate even without GPS.  

It not only analyzes, it fixes. At each waypoint, Overwatch can course correct the drone’s positioning to ensure it doesn’t drift. 

It uses AI to continuously learn overtime and at scale. Every time Overwatch is used, it generates sensor data for training. With machine learning capabilities, this means higher accuracy every time it’s used. And when more systems connect to Overwatch, the faster the training compounds. 

Best of all, it’s deployed as a mobile app. Overwatch can be downloaded onto mobile devices through Google Play Store (still in review right now), allowing users to access GPS-denied navigation without complex integration or new hardware. This  also means it’s not limited to drones and can extend into broader use cases.  

Tldr; Overwatch can turn a $500 drone into a high-grade system. It’s cheap, easy to deploy, and solves a problem worth hundreds of millions, if not billions.  

Very exciting product, and the investment case for me is fairly straightforward. But with that said, Sparc AI is only at the beginning of its commercialization journey. Here’s what I think it’ll take for it to grow long-term. 

Catalyst #1: Overwatch works as intended in real world scenarios 

This is the most important catalyst by far because it is the one that validates everything else - it’s proof of Overwatch working in real life. Right now, public knowledge of Overwatch’s capabilities mostly come from Sparc AI’s news releases. 

Fortunately, Sparc AI is making progress here. 

Last month, the company announced an order from a group operating closely with the UAE Ministry of Defence for Overwatch to be deployed for field evaluation.  

Earlier this month, Sparc AI then shared news that a leading Indian defence drone manufacturer would also be evaluating Overwatch. That is notable because this manufacturer supports the shift toward smaller drones intended for mass deployment and will become a strong ally for Sparc if they validate the tech. 

As you can see, these news stories are extremely recent. Right now, the market is still being asked to believe the technology works. Once trials begin generating real-world feedback from defence operators, that question starts moving from theory to evidence. 

Catalyst #2: Scaling from trials to closed deals 

Validation alone is not enough. Plenty of defence technologies work but never scale commercially. 

The next major growth driver is whether the current Overwatch trials convert into broader deployments and whether Sparc AI has the channels to distribute the product efficiently once they do. 

On that front, Sparc AI appears to be trying to solve both issues at once. In November 2025, the company signed a non-exclusive agreement with Precision Technic Defence Group, who will handle distribution across Australia, Europe and the US.  

Sparc AI also recently established an entity in the US to engage with the US defence market, which has the largest market share of drone demand and is allocating $33 billion for drone spend in the next few years.   

Catalyst #3: Expansion beyond defence into commercial and civilian use cases 

Defence is the most obvious proving ground, but I don’t think it is the full story. 

Looking deeper at Overwatch’s capabilities, I think you’d also see its potential outside of modern warfare. The company’s platform is described as being applicable across drones, unmanned ground vehicles, robotic systems, maritime vessels, etc. 

Low GPS capabilities, weak signal environments and navigation challenges exist in remote areas, mines, dense cities and many other environments. 

In fact, military demand in the GPS-denied market factors for just over half of total demand. 

On top of that, Sparc AI is aware of this and intends to tackle multiple market segments at once, as shown in its TAM research in its latest report.   

Sparc AI’s research shows some key numbers: 

• The market for Drone Navigation Systems is expected to grow at a 32% CAGR to 2030. 
  • Navigation Software specifically is expected to grow to $12.6B by 2034, the fastest growing sub-segment. 
• Commercial drone use is expected to grow 40% from 2022 to 2030 - and it’s expected that many of these will prefer GPS-denied navigation capabilities. 

Once the software is proven in one of the hardest environments possible, it becomes easier to imagine adoption in lower-stakes commercial settings. I wouldn’t make commercial expansion the core of the thesis today but I do think it is a legitimate second-wave growth driver if execution goes well. 

Valuation: $66 million market cap entering a $54 billion market 

Sparc AI is now up 270% year-to-date, but is still extremely small compared to the market it’s tackling. I’m not surprised - it’s just starting to execute its commercial strategy, and there are lots of unknowns around its tech. 

So here’s how I would value it. 

First off, Sparc AI has three revenue streams as outlined in its investor presentation. 

• Overwatch: annual fee per deployment of Overwatch. 
• Licensing fees: annual fee per drone and anchored to the value of the drone.  
• API consumption: priced per 1,000 solves; probably for when Overwatch uses Sparc’s AI capabilities to solve drift or any other issues. 

The only key numbers we have are: 

• $2,950 per drone licensing fee: the only revenue-related publicly available information that Sparc AI quoted the group based in UAE. Which I’d guess falls under revenue stream #2. 
• 95% gross profit margin: Sparc AI’s estimate based on the three revenue streams.  

Higher than the SaaS average, but not unbelievable considering the annual Overwatch deployment fee, economies of scale with a per-drone licensing fee, and API revenue.  

 Without knowing the numbers of revenue streams #1 and #3, I’ll anchor this valuation on just the licensing fee and the gross profit margin. 

Key assumptions:  

The US Department of War plans to build a drone force of 300,000 units by next year, India currently has around 40,000 registered drones, Australia had 3,100 in 2023 and plans to invest at least A$12 billion in new drones over the coming years.  

Assuming Sparc AI captures a very small share of this market, we model a scenario of 10,000 deployed devices in the next three years. This represents only a fraction of total available drones across these regions. 

Assuming ~$3,000 annual licensing fee per device. This is based on the company’s disclosed pricing of $2,950 per unit in its UAE-linked deployment, which provides a real-world benchmark for recurring revenue.  

Assuming a 30% operating margin (EBIT margin). While SPARC has indicated ~95% gross margins due to its software model, I’ve taken a more conservative view at the operating level to account for growth, R&D, and scaling costs. 

Using a 10% discount rate. Given the company is early-stage, operating in defence, and still proving product-market fit, I’ve applied a higher discount rate to reflect execution risk. 

P/E multiple of 20x. This aligns with typical benchmarks for high-growth software companies, particularly those with recurring revenue and strong margins. 

Price target of CA$5.25 per share: 

At 10,000 deployed devices, Sparc AI could generate ~$30M in revenue and ~$9M in EBIT. Applying a 20x multiple implies a valuation of ~$180M, or roughly CA$7.00 per share based on 26m shares outstanding. 

Discounted back over a 3-year period at a 10% rate, this suggests a present value of around CA$5.25 per share. For me, this means Sparc AI may have an upside of 104% based on its closing price of CA$2.57. 

To me, this calculation is rather conservative too given that the 10,000 units was only based on the number of government-owned drones. My calculation does not include an estimate for civil and commercially-owned units, which as mentioned, may be another key growth catalyst. Similarly, I only use one revenue stream (licensing fees) and gave a pretty conservative EBIT margin. 

Risks 

While you can see my stance with Sparc AI, as always there are risks to investing in an early-stage company such as this. Here are a few biggest ones: 

• Technology validation risk. Most of Overwatch’s capabilities are still being proven in real-world environments. Until we see consistent performance across deployments, there’s a risk the system doesn’t deliver the accuracy or reliability expected. 
• Trial-to-contract risk. Early traction is coming from trials, but defence procurement cycles are long and uncertain. There’s no guarantee these evaluations convert into large-scale, recurring deployments. Similarly, defence contracts tend to be all or nothing - either a company wins the entire contract or doesn't - which means higher risk.
• Adoption risk. The valuation hinges on scaling to thousands of deployed devices. If adoption is slower than expected, or limited to niche use cases, revenue may fall well short of projections. 
• Execution and competition risk. SPARC is still early-stage and will need to scale operations while competing against larger, better-funded defence and navigation players who could enter the space.  

Conclusion  

As I’ve outlined, Sparc AI is trying to solve a real and growing problem. Its approach is simple but powerful: instead of adding expensive hardware, it uses software to make existing systems significantly more capable. If that works in reality, it has the potential to become a scalable layer across drones and other autonomous systems.  

Importantly,  even a small share of deployed devices across a few regions is enough to support meaningful revenue and justify a higher valuation.  

That said, the story still comes down to execution. Over the next 12 to 24 months, the key questions will be whether the technology consistently works in the field and whether trials convert into real deployments. If both happen, the upside could be significant. If not, the valuation case weakens quickly. 

Additional information I’d like to share 

 Other qualitative areas of a company can also expand on an investment thesis. For this specific instance, that’s the belief that Sparc’s management team has in the company. 

Sparc AI CEO Anoosh Manzoori has led every financing round for the company, demonstrating his commitment to it, and has not taken a salary. He now controls 30% of the company.  

Not only that, the company’s directors have been buying up Sparc AI over the last few years, without a single one of them selling.  

This is to me a fantastic testament to their belief in Sparc AI.