SpiderOak has partnered with leading defense contractors including Northrop Grumman, Lockheed Martin, Raytheon, and Ball to test and tweak its software-based end-to-end cybersecurity and resiliency solutions for government, military, and commercial space operations. In June 2023, the company successfully demonstrated its OrbitSecure software in low Earth orbit on a Ball Aerospace payload aboard a Loft Orbital satellite. We interviewed SpiderOak’s Senior Vice President for Space, John Moberly, to find out how SpiderOak intends to deliver a disruption-tolerant space networking future.
Crispin Littlehales, Executive Editor, Satellite Evolution Group
Question: Please provide a summary of SpiderOak’s expertise and evolution in the space industry. What portion of the company’s efforts are targeted towards military users?
John Moberly: Just a few years ago, one of our existing customers challenged us with a surprising question: “Would SpiderOak be able to provide a similar, bullet proof solution for vendors and integrators who need end-to-end data protection for a space system?” The unique challenges, as compared to the others we had faced in the previous 12 years, were two-fold. First, satellite systems are intermittently connected networks and second all future architectures appeared to be on small, low-power devices connected in meshed networks. SpiderOak engineers are notorious for loving impossible challenges, so after working with potential customer experts in the space industry, we created OrbitSecure—a fully decentralized software platform for secure data orchestration (storage, transport, validation) in space.
By “space” I’m not just talking about satellites. Yes, we protect the data that enters the satellite, the data that leaves the satellite, and the data that belongs to the satellite. But we are also protecting the uplink, the downlink, the ground stations, the antennas, and the terminals. In addition to securing the data in transit, we secure the data at rest.
We are bringing a software capability that is secure, distributed, and fully decentralized. We can upload it to various elements already in orbit, and we can build out new capabilities. What sets us apart is that we can update the on-orbit satellites. That is critically important. If you are not able to update on-orbit satellites and be backwards compatible with them, you’re just creating your weakest link in space.
Right now, we are mostly targeting military and defense users. We just closed $16.4 million in an oversubscribed Series C round of fundraising. The OrbitSecure software with which we recently achieved flight heritage is already a minimum viable product. With this raise, we have the resources to further enhance our software and productize for the different space missions over the next year or two.
Question: The company has forged some interesting partnerships with TriSept, Raytheon, Lockheed Martin, and Ball Aerospace. What attracted these companies to SpiderOak and how are those partnerships unfolding?
John Moberly: Whether we’re working with a vendor, integrator, or solution provider, it always starts with a question: how can SpiderOak help enhance my existing offering to meet zero-trust requirements in space and win contracts? Given the challenges of space – size, weight, power, connectivity – only a decentralized and asynchronous solution will actually work.
We’ve been told we are the only solution with a fully decentralized system for data orchestration across all domains—ground, air, sea, and space—and these partners were looking for an accelerated, proven pathway to meeting true zero- trust. All these relationships started with rigorous internal testing of our solution in their development environments, including flat sats, to confirm our small footprint, performance, and of course, security.
We do have working relationships with the satellite Primes including Lockheed Martin, Ball Aerospace, Raytheon and some who are notoriously secretive. We are also collaborating with other developers. A good example is TriSept, which is developing a secure operating system called TSEL. For all the companies with whom we work, we always aim to be the valued mission partner that can solve secure and trusted data orchestration problems for them and their customers, typically government and military clients. So, those are true win-win partnerships in our eyes as well.
We have the OrbitSecure platform which is fully decentralized. It is built upon a foundation of zero-trust encryption and distributed ledger. It has role-based access control, identity management, and security. In addition, it has the peer-to-peer message queuing as well as peer-to-peer file sharing and storage. All of those together make up this baseline and then it’s easily modifiable to suit the various needs of our customers. We don’t get directly involved in sensor design or any of that, to us all those satellites are simply nodes in an autonomous network. The data going in and out of every memory stack is what we’re protecting right through all the access points that we mentioned earlier.
The big goal right now is to create a fully resilient zero-trust mesh network in space so that information can be passed ubiquitously through other systems and reach the intended user, decision maker, or warfighter on the ground. Our strategic partnership with Raytheon Technologies’ BBN division is all about developing and fielding a new generation of zero-trust security systems for satellite communications in proliferated low- Earth orbit (pLEO). We are combining OrbitSecure with Raytheon BBN’s Distributed, Disrupted, Disconnected, and Denied (D4) secure cloud solution to ensure resilience of mesh networks in contested environments.
Many companies are talking about doing distributed architectures and on the government side, the Space Development Agency (SDA) and a couple of other groups are moving aggressively into that, but the space sector is still in the early days. Even the Starlink constellation is not a fully meshed network.
SpiderOak makes a very lightweight software protocol, a very thin client if you will, that can be uploaded onto existing systems in space. That’s not easy. When you take into consideration orbital speeds and the vintage of some processors, you have a very limited amount of time to upload. On June 22, we were able to successfully deploy our OrbitSecure software and integrate it into Ball Aerospace’s hosted payload in LEO, making it the first time a zero-trust application has been performed in space. That was a significant achievement for us.
At present we are continuing to validate the entire baseline platform in the operational space environment. But then, as we tailor and modify it for specific use cases, we are also creating and proving multiple product lines.
Question: An increasing number of MilSatCom satellites are in orbit today with more scheduled to be launched. What portion of these systems are vulnerable and are there ways to make all of them cybersecure?
John Moberly: Securing satellites is, in many ways, much more difficult than terrestrial cybersecurity. But in some ways, it is safer. If we assume that the supply chain and the satellite are fully secure going into orbit, then the primary ways in are through the uplink and downlink. We can secure those with our role- based access control so that when an adversary or any malicious code that doesn’t meet that authority and permission tries to enter, it will be denied. That is the key part of what SpiderOak delivers at the intersection of space and cyber.
Question: How can security software be uploaded on-orbit and what are the challenges to making that option a widespread reality?
John Moberly: It starts with having a software-only solution – built in a memory-safe language like Rust — that can not only work on new hardware but legacy, flight-heritage hardware that many satellite systems continue to employ. But it’s not just having a small memory and storage footprint, our entire platfor is built around an asynchronous protocol that works over slow and intermittently connected networks. We’re able to upload just as we demonstrated on the Ball Aerospace payload which incorporated Ball’s Open Software System (BOSS) framework, which is made for swift data processing and modification of applications on-orbit. That was just the first demonstration, we have more planned on other satellites and ground systems. Our first was challenging in and of itself because the satellite was small, so it didn’t have big antennas and it didn’t have big link budgets.
Question: During any military mission, secure and trusted data must move from sensor to shooter/decision-maker while traveling through many nodes of variable trust. How can warfighters trust that the data hasn’t been tampered with during that entire chain of custody?
John Moberly: Since our platform is role-based access control by design, you must have the authority and permission to even get in. What’s more, our software has a distributed ledger which works much like the traditional blockchains used in cryptocurrency.
You can see the full data provenance from the sensor all the way down. This makes it easy to determine whether anything has been tampered with in transit throughout the chain of custody. That’s 90 percent of the battle right there.
Understanding that the warfighter who is in the battlefield needs to know immediately if the data is secure, we are developing a simple automated way for checking that the data provenance has been held throughout. We are envisioning using something like a red or green light as an indicator. We will also have a graphic user interface that enables monitoring of our software for both safety and some of the data provenance capabilities.
Question: Relying only on hardware encryption significantly increases latency while also causing a timely and complex re-keying process if there is any compromise across any node in the architecture. How can SpiderOak mitigate that particular risk?
John Moberly: Traditional authentication using just hardware involves going back to the central data center for authentication then back up to the satellite. SpiderOak’s software works with any encryption, and we can decrease the latency and increase the agility of the queuing system. Since our system is fully decentralized the message queue handles all the authentication back and forth in a peer-to-peer manner continuously. Many of our partners also use hardware encryption on each link of their system, whereas we provide encryption end-to-end, regardless of the number of hops from source to destination: this provides a significant latency and power advantage over traditional approaches.
At the same time, we’re changing the keys and doing what we call expert key management but without doing the encryption and without doing the ECU because we don’t want to get involved with the whole NSA certification process at this point.
Question: Can your platform be deployed to satellites in LEO, MEO, and GEO?
John Moberly: To us, it doesn’t really matter which orbits the satellites are in, or even how many different data paths need to be protected. Again, they are just nodes in an increasing complex network. However, we must be able to integrate into different types of hardware because of some challenging radiation environments that some processor hardware was designed for. Our software is also agnostic when it comes to encryption algorithms, we can use off the shelf commercial encryption or customer directed special purpose software – plug and play if you will. We’ve already worked with multiple types of processors and even have a project destined to land on the Moon. Although it was quite difficult to embed software into that operating system, we did it.
Question: Is it possible to stay a step ahead of hackers and other bad actors when it comes to cybersecurity?
John Moberly: It’s very difficult. We feel that prevention is 90 percent of the cure, and you can get that from our software through the authority and permission access controls which makes it incredibly difficult to infiltrate. Our customers are especially interested in our solution because it works well in the protection of completely autonomous systems, like those out there alone in space.
Question: How do you see things unfurling for SpiderOak in the next 1 to 5 years?
John Moberly: We have the resources to continue to evolve our initial set of products that are currently at the minimum viable product and proof of concept stages. Meanwhile, we are starting to sell those products off the shelf with enhancements and some operation maintenance such as sending updates. We will then start scaling up because as we solve more of the real operational challenges our customers encounter, we’ll take on even harder engineering problems, but always building from the decentralized, secure data transport platform. The goal is to create assets that will continue to matter well into the future.