MDA into Canadarm3 development as commercial space robotic needs merge

by Nathan Barker & Chris Gebhardt

From its first flight in November 1981, to the on-going evolution of the technology to take it from low Earth orbit out toward the Moon, the various Canadarms and their associated space robotic platforms have been an ever-present and critical element of human space exploration.

Now, as the need to develop a more autonomous robotic system for the Lunar Gateway arrives for MDA of Canada, the company simultaneously finds itself working to meet the growing commercial demand for space-based robotics. 

Across nearly 40 years of operations, the five Canadarms and three Orbiter Boom Sensor System (OBSS) arms for the Space Shuttle, as well as Canadarm2, Dextre (formally known as the Special Purpose Dexterous Manipulator), and the other elements of the Mobile Servicing System (MSS) for the International Space Station have amassed more than 3 million hours of engineering and operational support for various on-orbit activities.

“That body of knowledge, from an operation procedure perspective — and the things you have to anticipate, react to, plan for, deal with, have solutions for — is a tremendous asset as you try to extract that knowledge base and put it in an AI-based control environment,” enthused Mike Greenley, CEO of MDA in an interview with NASASpaceflight centering on the interaction between the company’s simultaneous development of Canadarm3 and its commercial market arms and systems. 

“It’s a tremendous asset to have at our disposal as we deal with this next generation of evolution.”

The largest element of that asset stems from the development of more autonomous uses of the arms.

The OBSS (left) rests in its cradle as Canadarm (right) is maneuvered for operations. (Credit: NASA).

With Shuttle’s Canadarm, crew members on the Flight Deck would manually control the arm, using their eyes as well as monitors feeding them views from cameras on the end for the arm for operations.  Over 30 years, the first generation of Canadarm provided ample opportunity for repetitive use and functionality of the arms — a key first step toward trust in automation.

As with Shuttle, the Station’s Canadarm2 started with crew-only operations in the early 2000s.  It was not until Dextre launched in 2008 that autonomy and a transition of control capabilities from the Station down to the ground began.

“So you’ve got this very, sort of, over decades evolutionary shift of building a solid hardware system that gets to be proven and then gets to be tweaked and modified that’s now operated for 20 plus years as a hardware system and a robotic solution on Space Station, but then a control environment that gets to shift from a Shuttle, to a space station, and then down to Earth.”

“It’s the same for us in life.  We don’t start with saying, I’m going to wrest control of my life to this machine.  You kind of try it out a little bit.  And through repeated action you kind of go, ‘Okay, I’m comfortable with this.’”

Mr. Greenley added: “I think that in a mission critical human spaceflight environment, the standards by which you go, ‘Yeah, I trust this,’ will be a little higher.  But the psychology will be very similar.”

Just as trust in the robotic system’s ability to work as designed when commanded was proven time and again on Shuttle, growing confidence in telerobotic controlling as well as automation of Dextre and Canadarm2 has developed in recent years.

In 2020, a major step forward in Station robotic autonomy began with the introduction of the MSS Application Computer, or MAC, a detailed overview of which can be found here.

But even though autonomy didn’t arrive until quite recently, the idea has been present since the first Canadarm for Shuttle was under development.

“The Canadarm designers and their counterparts at NASA envisioned automatic operation way back in the 70s,” said Cameron Ower, Chief Technology Officer, MDA.  “That highlights one of the barriers here in that it’s not only what the technology can do, but it’s what people trust.  It’s the whole thing of autonomy and autonomous cars.  That extends to robotics and space as well.”

While trust in automated systems — both Earth- and space-based — is increasing, it is not just comfort with automation that drives its need as a major part of Canadarm3’s design.

The highly constrained communication windows for robotic operations with the Lunar Gateway mandate, by design of the outpost itself, autonomy.  Controllers will not always be able to point a camera outside of Gateway at the worksite where Canadarm3 is performing operations.

This means the arm has to not only accurately sense and “see” its surroundings, but it also must maintain situational awareness at all times so it can re-plan certain tasks and moves on its own if needed.

While that second part has not yet been demonstrated with Canadarm systems at this time, the first part — vision — has.

“Just like for us human beings, we’ve got a plan in our minds, but we do final alignment with tactile or appropriate sets of sensors, or eyesight,” said Mr. Ower.  “That’s the evolution now: to be smarter you have to be able to sense the world in a better way.”

Canadarm2 reaches out and grabs the arriving HTV-7 resupply vehicle from Japan. Dextre can be seen in the upper right . (Credit: NASA)

Part of Canadarm3’s vision will incorporate similar, though not identical, systems employed on the OBSS arms for the Shuttle program, which were tasked with “looking” at the Shuttle’s wings and Thermal Protection System to search for and assess — in highly detailed surveys — any damage to the vehicle.

The OBSS, far more than Canadarm and Canadarm2 at the time, had to very accurately relay distance and situational awareness data to the crew in the Shuttle’s Flight Deck, as the areas the OBSS had to be maneuvered to were not visible to crew.

Mr. Ower related, “To gain greater autonomy, often people jump to the ‘thinking’ part of it.  But you can’t actually do that thinking unless you have a good assessment of the environment: where things are, how they’re oriented, or maybe even recognizing objects.”

Likewise, Canadarm3 will have to contend with lunar dust brought to the Gateway by landers returning crews either directly to the outpost or to its vicinity, a different radiation environment than its counterparts have endured, and longer-duration exposure to temperature extremes.

To meet the challenge presented to Canadarm3, and to further evolve from Canadarm2 and Dextre, Mr. Greenley sees three key areas that come into play: technology base, R&D projects, and MDA’s internal development and solutions for the commercial market.

An immense project like Canadarm3 will bring those three together.  As Mr. Greenley said, “We can combine the evolution of government programs with the things that have evolved from ongoing R&D and collaborative R&D to our own commercial solutions.  They can now get merged together to create the solution for the next generation.”

The simultaneous development of Canadarm3 and on-orbit robotic services for the commercial market is an overlap that has not happened before, as “spin offs” or commercial applications for Canadarm, the OBSS, Canadarm2, and Dextre were not truly considered during those development cycles.

“I was involved in the original Canadarm2 and Dextre design and development,” said Mr. Ower, “And there was no parallel interest in doing these things commercially.  This is what’s so exciting about the Gateway project right now.  Unlike those previous two generations, there’s an active interest to modify things in space, to actually build initial infrastructure in space.”

Mr. Greenley agreed: “It’s a very different lens.  We’ll be delivering Canadarm3 at the same time our commercial line of space robotics will be delivered to other businesses in collaboration with them for on-orbit servicing, debris removal, or assembly applications.  And so it’s a much richer technology environment with parallel, interacting technology roadmaps that we get to take advantage of.”

This interaction was seen the second week in January 2021 at a virtual Industry Day event surrounding Canadarm3 — which saw over 400 companies learn more about the requirements of the program and how their companies might assist.

“You’re opening up this dialog with industry and you’re building and focusing these conversations so that as you’re shaping your requirements and designs, you’re also shaping your conversations with your supply base during that period of time to make sure your relationships are ahead of your need so that you’ve got the right things that you need at the right time to hold the project on schedule,” noted Mr. Greenley.

An overview of Dextre. (Credit: CSA and MDA)

Mr. Ower added that “One of the interesting things of having these Industry Days is that it brings out organizations or institutions that have ideas or technologies that may be brought to bear on the problem immediately or they may be part of the evolution of the system as well.”

That second part is crucial, as MDA will likely be presented with ideas from industry that would prove exceptional for the Canadarm3 but that aren’t technologically mature enough to meet the project’s launch target between 2026 and 2027.

“The technologies that we’re working with right now as we get into the formal requirements phase [of the project], we need to be drawing on technologies that are already really at about the TRL 4 level or above,” said Mr. Ower.

“We may learn of technologies that may not be ready for first flight, but they may be things that feed into later developments of Canadarm3 or the related things that Mike was talking about in relation to immediate attempts to commercialize or maybe solve a slightly different problem.”

Crucially, though, the commercialization and commercial applications of the hardware and software systems developed by MDA for their robotics systems have direct impacts to everyday life on Earth. 

This includes the economic and job benefits of such programs like Canadarm3.

“One of the new things that I think is an important part of the whole spinoff conversation and the benefit to society and the world from doing this space stuff is that, the way that space is growing now, there’s a lot more space stuff,” said Mr. Greenley.  “It’s becoming an economy all on its own.”

Artist’s concept of Canadarm3, Canada’s smart robotic system, located on the exterior of the Gateway. (Credit: Canadian Space Agency, NASA)

Beyond that, too, are the life-saving implications and implementations of the technology.

Canadarm’s control systems for the Shuttle program have fed directly into the design and use of medical robots for brain surgery, where minute, precise, trusted-to-happen-when-commanded movements are crucial.

Current applications from Canadarm2 and Dextre are used for breast cancer diagnosis and are being developed for breast cancer surgery as well.

The newer, commercial robotic technology under development by MDA for space-based applications is also being actively considered as part of the next generation of medical robots that could perform surgeries on humans in remote locations or on those who lack access to a surgeon.

“I think that the level of autonomy that will come from Canadarm3 on Gateway introduces entirely new opportunities, especially for countries like Canada, where you have a lot of remote operations,” says Mr. Greenley.  “Whether it’s under the water, or in the north, or in rural areas, we should see legitimate conversations for expansion of those technology transfer opportunities.”

Development of Canadarm3 began with trade studies conducted from 2014 to 2020, with MDA formally receiving the contract and authorization to build the arm as one of Canada’s multiple contributions to the Lunar Gateway project and NASA’s internationally-backed and supported Artemis program.

Additional features chronicling the development, build, and launch of Canadarm3 will follow.

(Lead image: Canadarm3 on the Lunar Gateway. Credit: CSA)

Related Articles