SkyfireAI merger puts spotlight on autonomous swarm drones in public safety sector

Skyfire Consulting is no more. Instead, the Georgia-based consulting firm that focused on the drone first responder (DFR) industry, has evolved into a new company. It’s called SkyfireAI.

That’s because Echelon AI, an autonomous and swarming Artificial Intelligence company for Uncrewed Aerial Systems (UAS), acquired Skyfire. It’s a merger of source, blending their expertise and their names.

Here’s everything you need to know about SkyfireAI, including how it got to where it’s at today, what’s in store for the future, and what this means for the future of drone swarm technology:

A history of Skyfire Consulting

Up until now, Skyfire was known for supporting first responders who used drones from start to finish. That included offering drone training services such as through its Skyfire Drone Academy and by offering workshops at drone conferences worldwide. It’s also helped clients navigate regulations such as applying for various certificates of authorization (COA) and specialized drone waivers.

It’s made its own hardware, including an American-made drone for public safety called the SF2. The mid-size drone (it’s 22-inches, so just under two feet) is payload-agnostic. That means it can carry all sorts of objects that a first responder might want, whether that’s a visual or thermal camera, or perhaps something like a speaker or life raft.

The company first launched in 2014. CEO and founder Matt Sloane has a background in first response given his work covering breaking news at CNN. He also has a background as an EMT. After bringing on other experts in drones, first response and business, Sloane launched Skyfire with a mission “to revolutionize public safety.”

Since then, it’s clients have included big names like Pacific Gas & Electric (PG&E). PG&E specifically sought out Skyfire to help it acquire a BVLOS waiver so it could fly large sections of its critical infrastructure within high fire-threat districts across the entire state. Also in California, Skyfire worked with the Fremont City Council to help create a unique, shared drone first responder program across the city’s fire and police departments.

The Echelon AI acquisition

Arlington, Virginia-based autonomous swarm drone company Echelon AI announced it would acquire Skyfire Consulting in September 2024. With that formed the new company with a name blending the two: SkyfireAI.

So what does the new company, SkyfireAI, do? According to LinkedIn, it’s a painfully-clunky mess of buzzwords. Yes, their company description states: “Next-gen autonomous and swarming AI SaaS platform for drone systems in multiple domains.”

In human-speak? SkyfireAI will deploy their smarter technology to help first responders use drones in a way that requires less effort from human pilots. It’ll focus on clients from fields including law enforcement, fire and EMS, as well as defense applications such as border and perimeter protection.

Today we see incredible use cases of first responders using drones. Examples include stopping high-speed car chases (which pose dangerous crash risks) and instead following the suspect with a drone. After all, between 2018 and 2022, 40% of chases resulted in a collision. What’s more, 17% of chases resulted in an injury to a suspect driver, police officer, or bystander, according to the California Highway Patrol. 

But use cases like an officer flying a drone to track a car theft suspect still largely require manual effort. That can include physically taking the drone out of its case, setting it off to fly, actually flying it and manually sorting through all its data gathered through cameras and other sensors.

Given the work involved, many first response teams have to have separate teams, which can be costly and inefficient. With SkyfireAI, artificial intelligence-fueled software could cut back on scale issues. Instead, it could enable the rise of defense applications for drones around the world. SkyfireAI promises to offer more turnkey drone solutions, with an emphasis on the right mission control software.

Eric Malawer, Echelon co-founder and COO, said in a prepared statement that drone hardware has become “increasingly commoditized.” That’s perhaps a reference to how just a few players tend to dominate the hardware space.

“Real differentiation lies in the AI mission control software and its associated technology stack,” Malawer said.

That software leverages generative AI (that’s something like ChatGPT, which can create something new for the user). SkyfireAI also has its own proprietary mission control platform. That means eaning users can use Skyfire’s software to handle most aspects of mission operation (as opposed to you just manually flying a drone yourself). Robust software also becomes especially critical in situations where one operator might handle multiple drones at once.

In short, rather than just supporting public safety agencies get off the ground as Skyfire once did, the new SkyfireAI will offer what it promises to be a robust, AI-powered software program to support its flights.

Swarm drone technology is key

With the acquisition comes a newfound focus on swarm drone capabilities.

Swarm drones are exactly that. Swarms are multiple drones that coordinate their actions to accomplish complex tasks that would be impossible for a single drone. In the past, it was one drone to one operator. With swarm drones, multiple aircraft our coordinated across the same network.

We see this with drone light shows. One coordinated system has the drones work together to fly in certain spots in the sky. That system lights up the drones at certain times, generating designs that only make sense when multiple drones fly in sync with each other.

Yes, that giant Mickey Mouse head displayed above is one giant drone swarm. (That was for a drone show at Disneyland Paris to celebrate Bastille Day. At the time, that ‘swarm’ broke a Guinness World Record for “Largest aerial display of a fictional character formed by multirotors / drones.” Though, Sky Elements since broke that record.)

Of course, drone light shows are just the beginning, as evidenced by this tweet:

This isn’t just a light show. It’s a demonstration of futuristic military power. You are looking at an unarmed drone swarm. https://t.co/ggDPtoZyCt

— Marko Jukic (@mmjukic) September 12, 2024

We’ll see swarm drone technology in search and rescue, environmental monitoring, and disaster relief. As this poster suggested, we’ll also see it in military applications.

Types of swarm drones

Now there is no single type of drone swarm. Instead, swarm drones function in a variety of ways, such as:

1. Centralized control schemes: With this, a single control point (like a ground control station) processes information regarding all the drones, and then near-instantly makes decisions and issues commands to the other drones based on that information. In this scenario, each individual drone isn’t necessarily aware of the one next to it. It reacts purely because of a command from the ground control station.
2. Decentralized control schemes: On the other hand, swarm drones operating in a decentralized control scheme could make a decision around the drones flying around it. These drones — though they might be limited based on predefined “rules” — still respond to other drones in the area through information gathered via onboard sensors, like cameras.
3. Distributed control schemes: In this system, each drone is autonomous but is able to share real-time information with drones around it. From there, it can make its own decisions based on that shared information.

The team over at drone analytics and consulting group Drone Industry Insights spelled this out in a blog post and graphic (shared below) all about swarm drones.

Swarm drone software matters

With swarm drones, it’s the software that matters far more than the hardware. Sure, drone aircraft quality is constantly improving with features like longer flight times or ability to carry bigger payloads. But considering that you can buy high-quality drones for under $500 (hey, the impressive DJI Neo is less than $200), it’s the next step of software that moves the needle.

As AI technology continues to improve and become more accessible, expect to see more swarm drone software, like what SkyfireAI is building, come into being. From a relatively basic standpoint, software like Verge Aero Composer makes your own drone shows for you. You simply input the colors and designs you want (and in which order and at what time). From there, the Verge Aero Composer software programs your drone swarm, aka drone light show, for you.

Of course, more robust operations necessitate more robust software. Techniques like artificial neural networks (ANNs) and deep reinforcement learning (DRL) allow drones to not just fly in sync, but also make smarter decisions that are optimized as they learn more about their environments.

Skyfire’s new emphasis on AI is evidence of a broader trending in the drone industry. Earning your Part 107 certificate and being a talented pilot il likely not the key to wild success in the future. It’s knowledge of robust software. It’s embracing AI — and working with it to allow it to do its thing. The future of drones is not one pilot to one drone. It’s many.

Did you find this article helpful in explaining swarm drones? If so, please consider making a donation to The Drone Girl. You might make a one-time donation and buy me lunch, or consider a recurring donation to ensure original stories and critical analysis continues!

The post SkyfireAI merger puts spotlight on autonomous swarm drones in public safety sector appeared first on The Drone Girl.