OT Interviews: Karen Holland, XCAM CEO

A press release did it’s work earlier this year: it caught our attention. XCAM was developing the world’s first particulate detector that would go into a rocket. It seemed odd that after decades of regular launches into space, only now are detectors being made that can tell us about what’s happening inside. So, we jumped at the chance to interview XCAM CEO Karen Holland and learn about what’s happening in the world of designing and building sensors for the space industry.

OT: XCAM’s roots go back to to the Leicester Space Center in the 1990s. How have things changed for the company since then with the rise of New Space?

Karen Holland: Although we were a spin out from the Leicester Space Center originally, we didn’t actually work in space until about ten years ago. We made equipment that space scientists used for the characterization of detectors on the ground, but primarily for the first 20years or so of our history, we made custom camera systems for use on big vacuum systems, big science experiments and synchrotrons and things like that. Increasingly over the last 12 or 15 years, we’ve made cameras that are used in the semiconductor manufacturing process. What we were good at is making cameras that are super, super clean and quite often for EUV or x-ray wavelengths.

A tipping point for us was in 2014, the UK Space Agency launched their first cube satellite with an imager on it that we designed and built in six months flat for the Open University, who were responsible for building the experiment. That was the beginning of our our work in space, really, it was that first imager system.

I think that New Space type activity has brought space down into the reach of businesses like us. Because when space was about using radiation hardened, or rad tolerant components that cost tens of thousands of pounds, small companies like us just couldn’t get into that that sort of business. Whereas when we can use COTS, albeit with a lot of care and thought behind the design, it brings it all within reach of a small business like us. So I think the knowledge we had in making clean vacuum systems helped us a lot. And then our CTO, his main career for many years has been that he specializes in understanding radiation damage in imaging detectors. He and his research group do that work for NASA and ESA and so on. Those two areas combined meant that we were quite well placed for exploiting this growth in New Space activity.

OT: Do you find that using COTS with commercial entities that never really had an opportunity to work in space or to found a use case is easier? Or, do you have to guide them a little bit as opposed to NASA or SpaceX saying, hey, we need this; it has these characteristics. And you’re already talking to a space professional. Is there more education these days?

Karen Holland: I think so. We’re working on a project for ESA at the moment, this is the particulate monitor that we’re developing for them for a Vega C launcher. It’s a relatively low cost project, so there are a lot of areas where we’re explaining that we cannot follow the standards that ESA would specify for large budget projects, and that we must adopt a new approach for low budget instruments. For as long as we’re not presenting a risk to the rocket and we can prove that there’s a low risk of failure, they seem to be willing to adapt.

We’re also working on this NASA Wind Cube project. NASA seem to have a scaled approach to the management of space projects where the requirements are much lower for relatively low budget projects. That’s quite refreshing as well. In that case, it’s about working with the team at the US university and agreeing on what we need to do to maximize reliability and performance within the budget and timescale envelope.

OT: Interesting. So it’s almost like the opposite instead of the educating the commercial user as to what is possible, you’re having to show the majors what can be done instead.

Karen Holland: Historically, ESA and NASA have worked on these big expensive missions. And they’ve seen what the commercial world is doing in terms of putting up satellites with COTS, with commercial parts at very low cost. It’s a really exciting period of change where they’re seeing these benefits and they want to start to benefit from them. But at the same time, if something is lower cost, then it will usually be of lower reliability or there will be trade offs and it may be not necessarily lower reliability, but you might need to reset your electronics every now and again or something like that. So I think they’re looking to benefit as far as they can from all the knowledge that’s being gained in this area.

OT: Joshua Western over at Space Forge was saying that they went with automotive parts for some things because the cost of getting those type accepted was cheaper than the cost of the already space rated parts.

Karen Holland: Automotive parts already go through a very high level of screening, so they’re pretty good parts to use.

OT: But what you’re talking about is actually going back to ESA and NASA and saying, “hey, look, there’s a better there’s a more cost effective way that doesn’t necessarily affect performance.”

Karen Holland: It depends on your contractual relationship with them, but in general, if the budget is low, then there is a need to accept that. With the particulate monitor that we’re making for ESA, the instrument doesn’t have to last very long, and it’s possible and sometimes necessary to make some compromises. So I think these relationships are always a dialogue between the customer and the space agency anyway. And perhaps they’re becoming more open to looking at some of these methods that the commercial sector is developing.

OT: There’s a lot of attention to artificial intelligence and machine learning with image processing these days. But what about hardware? People are focused on the software side. Where do you see the hardware technology going in the next few years? Is it again, more budget consciousness, or is there a new development that’s sort of on the horizon that people are looking at and wondering how it’s going to actually be wrought out, or where’s the hardware going?

Karen Holland: AI and ML are in our thoughts related to all new developments. That’s because the bottleneck in space is getting data down to the ground and you need AI to be making decisions up there so that you’re not sending data down to the ground unless it’s useful. You can have decisions made in space, so we’re going in two different directions. In some senses we’re looking at including AI capable FPGAs within our hardware to do processing on board. And then we’re also partnering with organizations who are more into the software side and the processing side to make sure that our some of the newer camera systems that we’re developing will immediately be able to pair with their AI and their processing systems, because it feels like this is a really important area going forwards. That’ll be at the heart of a lot of what we’re doing. The particulate fallout monitor that we produce currently uses a software algorithm to identify contamination that has fallen on a sensor and to extract those those shapes and sizes. But contamination can be really complex. It can be a little speck of something. It can be a curly fiber that sits partially on the surface and partially off the surface. And using AI ultimately is going to be the only way to go to do that analysis. We also don’t want to be doing that processing on the ground. It has to be done in space and we just send the data back that’s of interest. So I think it’s going to be really important in the future for us.

It’ll either be on the imager controller board, where there’ll be a capability to take that image and do processing on it or it may be a separate processor board. It can be either one of those two.

OT: Regarding the particulate contamination monitor, it just it struck me that we’ve been sending up rockets for, you know, for decades now, and they’re only now finally getting around to figuring out how to do this. And then also the first space weather observation platform. But what else is there? With this change in imaging, are people asking what can we observe that we couldn’t before? Or are there different or old questions being thought of in new ways? Do you see your people looking at your products and capabilities and saying, ‘Oh, hey, we could do this too. Why don’t we try it? Well, how’s that working out?”

Karen Holland: I think we’re seeing both sides. I think there’s a project that really excited me that we were talking to one of the organizations about, and I don’t think I can talk in detail about the project, but it was some science that was essentially, well, it’s difficult to know how to describe it. Some of the really big x-ray telescopes that have been sent up by organizations over the years are showing discrepancies between some of the measurements that are being done from different instruments. And there was a lovely little CubeSat project that would have been relatively low cost that couldn’t have been done five, five years ago, which could go up to check which answers are right. You know, which one is the right answer or what’s going on.

That was one of my favorite ones because it wouldn’t have happened five years ago. And it was such a low cost thing to check some of the science that was coming out of these really big telescopes. That’s one of the most exciting projects that we’ve come across.

A lot of cube satellites are used for technology demonstration, and that’s quite exciting as well. There’s a system we’ve just delivered to Penn State University, which is going to be launched on a sounding rocket, and it’ll be in space for about 200 seconds, maybe collect a thousand photons and it’s testing out a new type of diffraction grating that’s never been tested in space before and some new silicon x-ray mirrors. To use these types of technology to accelerate the development of other technologies for space is a big theme in science like the Wind Cube project and we’ve had other organizations interested in similar space weather projects.

The contamination project is just something that hasn’t ever been done before. And I think years ago, when everything was made of cleaned machined metal, perhaps it might not have been so much of an issue, but with 3D printed parts and lots of fibers being used now, Nextel and carbon fibers, I think there are some suspicions that there may be some contamination incidents occurring and yet nobody has ever measured what happens during launch whether contamination is falling on telescope mirrors that you’re going to launch into space. There may be other areas in the future that can be monitored, so there’s a lot of exciting stuff happening at the moment, which is both new science experiments and old questions.

OT: So what happened with the x-ray project? Did that get shelved or set aside or was somebody afraid of losing a bar bet?

Karen Holland: No, that one didn’t get funded. A bid was put in by a team. It didn’t get funded, but it may get resurrected at some point, perhaps.

OT: How about licensing? Do you see a need for for streamlining or is it for you all fairly cut and dried these days? What what can you tell me?

Karen Holland: We don’t tend to license because we’re basically an engineering business and we have a lot of know-how and expertise associated with how to design complex camera systems, and that’s mechanical and thermal design, loads of electronics and embedded software and firmware and so on. But a lot of what we’re doing is not very easy to protect. So we tend to get quite nervous about licensing if we can’t protect that knowledge.

We have a couple of patents that cover some aspects of the particulate monitor and where we have a patent, then maybe it’s possible to license. But historically we haven’t licensed because we can’t protect that know-how and expertise that we have. We tend to just try and protect it by not sharing.
We have a couple of patents that cover some aspects of the particulate monitor and where we have a patent, then maybe it’s possible to license. But historically we haven’t licensed because we can’t protect that know-how and expertise that we have. We tend to just try and protect it by not sharing.

Note: Orbital Today would like to thank Karen Holland for taking the time to share her insights with us.