Today, we’re kicking off 2021 with an industry-first achievement. SparkCognition recently announced a new collaboration with SkyGrid (a Boeing, SparkCognition Company) that makes it the first airspace management system to deploy AI-powered cybersecurity on drones to protect them from zero-day threats including ransomware, viruses, malware, and more. This is critical news as more and more drones are expected to take flight in the coming years spanning a wide range of civilian and commercial use cases. Below, you can get a first look at SparkCognition demonstrating its DeepArmor® cybersecurity product on a drone out in the field.
Based on the latest data, the Federal Aviation Administration (FAA) predicts the commercial drone market will triple by 2023, hitting 835,000 aircraft and far exceeding the FAA’s initial projections. In 2017, futurist Thomas Frey predicted that there will be one billion drones sharing the skies. The significant growth in numbers and complexity is incredibly exciting, however, this presents a major challenge in protecting aircraft systems from remotely being attacked by zero-day cybersecurity threats.
With the rise of communication between people and devices and the rise of computing performance, aircraft such as drones are not immune to cybersecurity risks that have been critical issues for other industries. When you think about what a drone really is today, it’s essentially a small computer that can fly, and with increasing drone activity, you have a network of small computers that are vulnerable to remote hackers. In an emerging drone environment, an AI-based cybersecurity approach is critical to more accurately detect and prevent cyber attacks from executing on a drone.
Integrated with SkyGrid’s airspace management system, AerialOSTM, SparkCognition’s DeepArmor cybersecurity product provides more advanced airspace security than traditional anti-malware that are reliant on signatures of known threats. Instead of using known signatures, heuristics, or rules-based approaches to detect security threats, the DeepArmor product uses patented machine learning technology and a layered protection strategy (known as Defense in Depth) to protect a drone’s endpoints. Not only can the DeepArmor product protect drones from known threats, but its machine-learning detection engine uses advanced classification algorithms to predict and prevent zero-day attacks without frequent updates, greatly enhancing protection.
A few key points to note from deploying the DeepArmor product on drones:
- The SparkCognition engineering and labs teams were able to streamline the DeepArmor product to run on the resource-constrained environment of the drone. Today’s drones don’t have a lot of spare computing power compared to your laptop, so a lightweight solution that minimizes performance impact is essential.
- Your network is only as strong as its weakest link, and if drones are part of your network and are out there in the world without onboard cybersecurity, you risk getting hacked. Case in point: in 2011, adversaries hijacked a US RQ-170 Sentinel spy drone. Electronic warfare experts were able to cut off the drone’s communications links and reconfigure its GPS coordinates to trick it into landing in enemy territory.
- The DeepArmor product can operate and protect drones autonomously without any human involvement, where it also supports a connected state for remote remediation and management of the drone.
Where do we go from here? SparkCognition CTO Sridhar Sudarsan stated, “We imagine a universe where you have hundreds of thousands, potentially millions of drones all being managed with the SkyGrid platform, managed within specific missions by region, by enterprise, by individuals, fully protected with DeepArmor.” What a beautiful world, indeed.
