Inventors on the frontiers of space and robotics show off their gadgets and tell the stories behind them

Four founders of companies on the cutting edge of technology gathered this week at a conference in Seattle to show and tell about the hardware at the heart of their businesses. And like any good show, their talks were all about strategy and gadgetry.

For example, consider the Seattle-based laser-powered weed zapper Carbon Robotics. The LaserWeeder system uses optical sensors and artificial intelligence to identify and target weeds between plants as the robotic machine is pulled through the field.

Carbon Robotics founder and CEO Paul Mikesell held up one of the LaserWeeder scanners during Monday’s DeepTech session at K&L Gates’ downtown office.

“We set it up so that this camera can see exactly what a laser shot like this is going to hit, and every time we turn on that laser, the same pixel area on the camera will explode and explode,” he said. “This machine reminds me of a lot of science and technology that we had to deal with, but also, a lot of pain that went into this thing.”

@media (max-width: 600px) { aside.callout { float: none !important; width max:100% !important; margin-left:0 !important; margin-right:0 !important; } aside.callout .callout-img { display:none !important; }}

The company’s engineers had to figure out how to precisely target the weed based on the image that was distorted by the camera’s viewing angle. “It’s a very good practice to do that the right way, and once we get it right, we just fire them every time,” Mikesell said.

Mikesell said he is often asked about his strategy for selling LaserWeeders through plastic-implement dealerships. “We decided to direct every time, every time,” he said. “And we have a global team of sales and service support. That is, we maintain customer relationships. We know what the products are used for, how well they work, what their challenges are. And customers know how to contact us directly instead of going through the sales floor.”

He looks forward to the day when artificial intelligence will speed up the hardware design process. “I’m surprised by the lack of an AI tool out there, but I think it’s also because, you know, software engineers have written software that does AI, so they’re very comfortable with it,” Mikesell said.

“We actually hooked Claude up to an oscilloscope and got him to generate the firmware that proved what we needed,” he said. “So I think that will continue to come.”

@media (max-width: 600px) { aside.callout { float: none !important; width max:100% !important; margin-left:0 !important; margin-right:0 !important; } aside.callout .callout-img { display:none !important; }}

Starfish Space Austin Link founder’s show-and-tell gadget was too big to take off the table. It was an engineering model of Starfish’s Otter Pup spacecraft, one of which is currently undergoing orbital docking satellite testing.

“We actually did a fun test over the weekend, which I can’t tell you about yet,” said Link.

The Otter Pup was designed to demonstrate the technology to be used on the full-scale Otter Starfish spacecraft to test or guide other satellites into orbit. “People have done this before, but every time we’ve done it, it’s really expensive,” Link said. “You look at the Northrop Grumman satellite that does the same job. They made $65 million by extending the life of the satellite. It cost $400 million to do.”

Starfish aims to use innovations in computer vision and robotics to make satellite docking more affordable. That means the Tukwila, Wash.-based startup. we must do more with less.

“This satellite has only one engine, and the energy generated by this satellite is the same as a housefly sitting in your hand,” Link said. “It’s a small force, so you have to use it thoughtfully over time. You have to predict what’s going to happen with the physics and then eventually you get together. And that’s our biggest challenge as a company, not just with the demonstration satellite, but ultimately with our full-size Otter.”

@media (max-width: 600px) { aside.callout { float: none !important; width max:100% !important; margin-left:0 !important; margin-right:0 !important; } aside.callout .callout-img { display:none !important; }}

Overland AI builds autonomous off-road vehicles, including a 3,000-pound tactical vehicle that can transport cargo, drones or even anti-drone weapon systems for combat units.

“It’s not very easy to find one in this room, but I wish I had brought one,” said Byron Boots, founder and CEO of the Seattle-based startup. “Instead, what I did was rip the sensor pod off one of these cars.”

The sensor pod is equipped with stereo cameras and a ranging lidar system, all connected to an onboard computer. “This comes from something called our SPARK Kit, which allows you to take any car and make it autonomous,” Boots said. “It actually hangs over the head where someone sits in a car like this.”

Although “AI” is part of the company’s name, Overland AI’s focus has expanded from writing software to building hardware as well.

“In order to go faster, we decided we had to do it ourselves and have this integrated vertical stack,” said Boots, a professor of machine learning and robotics at the University of Washington and CEO. “If you do that, you can give the user this robot system with an independent stack, and they can start using it. You don’t have to wait for someone else to put it together with you.”

@media (max-width: 600px) { aside.callout { float: none !important; width max:100% !important; margin-left:0 !important; margin-right:0 !important; } aside.callout .callout-img { display:none !important; }}

Redmond, Wash.-based Starcloud made its mark at the intersection of AI and space operations last year when it became the first company to train a large-scale language model in Earth orbit. For this week’s show, the company’s founder and chief technology officer, Ezra Feilden, brought an Nvidia H100 GPU – the same type of AI chip that was used for last year’s local display.

“It’s very high power density. These GPUs are designed to stay and have a good easy life inside the data center. They were not designed to be strapped to a rocket and launched into space, then run for five years without any maintenance or TLC,” said Feilden. “So, that’s part of what we do at Starcloud. We harden GPUs and other IT components to survive launch, and then be able to operate continuously in space without mechanical intervention.”

Orbital data centers are attracting a growing wave of buzz because they can address some of the biggest problems created by the rapidly growing hunger for AI data processing capacity: for example, limitations on available electrical power, and concerns about land and water use.

Processing data in space brings its own challenges, however. How many solar powered satellites will be needed to handle the load? How will they be connected? And how will they manage to get rid of the waste heat generated by all those high-powered AI chips? Feilden and his colleagues at Starcloud are working to address those challenges.

Feilden said his company is ramping up operations at the new facility in Woodinville. “We will be using a number of satellites that we build at that center in the next few years, going up to thousands of satellites a year, which is the number we need to hit to have a real impact on the terrestrial data center industry with what we do,” he said.

Starcloud isn’t the only company with big ambitions for orbital data centers. SpaceX, days away from the world’s largest public offering, plans to put a million data center satellites into orbit. But SpaceX can be as much a partner as a competitor. Last month, Starcloud struck a deal to use SpaceX’s Starlink mini laser terminals on its satellites. And Starcloud relies on new launch vehicles, including SpaceX’s Starship mega-rocket, to lower the cost of putting satellites into orbit.

“We strongly believe that will happen soon,” Feilden said. “This is the decade, really.”

Austin Link of Starfish Space was intrigued by that idea. He noted that Feilden and his colleagues are working on a business model that assumes that startup costs will be significantly lower. “We think startup costs are what they are today, and we don’t make any changes to our models when we design products,” says Link. “It’s a really interesting difference.”

The program of events for Deep Tech Week Seattle continues until Friday.