Neuralink’s Surgical Robot: The Designer’s Story Behind the Innovation

Afshin Mehin has become the go-to designer for companies working on devices that aim to tap into or modulate the brain. The creative agency he founded, San Francisco–based Card79, has worked with “, records and transmits brain activity with the goal of enabling paralyzed people to control a computer.

Mehin worked with Neuralink to design the external parts of this system—the installation robot and also a wearable that would sit behind the ear and transfer data and power to an implanted wireless receiver. This device, which looked like a sleek, white hearing aid, was an early prototype. Neuralink unveiled a ” in Alzheimer’s patients, and Science Corp., a startup founded by former Neuralink president Max Hodak that is developing a visual prosthesis for the blind. More recently, Mehin has been tapped to help design an “

The Zero Byte spoke with Mehin about his approach to designing brain devices and the future of neurotech. This interview has been edited for length and clarity.

The Zero Byte: How did you end up working with Neuralink?

Afshin Mehin: We’ve been working in the space of designing wearables for a decade now, almost a decade and a half. We worked with a company in Vancouver called Recon Instruments that was eventually acquired by Intel. The work we did with them was around the time Google Glass was being developed. These guys [at Recon Instruments] were doing a pair of sports sunglasses for running and cycling and it was all about understanding how to put technology around your head. We had to add an extra processor and a camera all the way up at the front of the glasses and so we had to get 3D scans of your head and databases of people’s heads to understand the anatomy of a person’s skull.

We started working with Neuralink in late 2018, early 2019. When they did their launch video, they showed the design we did. We met with them about six months before that and were introduced to them because of that background we had in designing for the anatomy of the head.

How do you design a brain-computer interface with user experience in mind when you can’t test it out yourself?

A lot of it was based on secondary research, understanding the pain points of what these individuals are going through, with the focus being on people with paralysis. We wanted to empower them to do things they weren’t able to previously do. In terms of the actual product design, we were thinking about the ways that either they or a caregiver could easily be able to manage the system.

,

As a designer, what safety considerations did you have to think about with the Neuralink device?

The primary safety considerations weren’t so much on the device but on the robot. We had a small role to play, which was to psychologically transform their first-generation robot, which was exposed steel—you could argue it looked pretty ominous—to something that was a little bit more approachable and ready for clinical trials.

We worked with Neuralink’s engineering team to try and design facias—covers or cladding—for the outside of the robot, to start to give it a bit more of a visual language that was simple, approachable, and something that you can imagine people not being intimidated by. In that process, we were starting to introduce a lot more elements of design, and the safety concern wasn’t so much for the patients, it was for the operators.

We thought about things like pinch points. You don’t want people crushing their hands while they’re operating the system. That’s Robotics 101. It’s what every designer who’s designing robots has to think about. These machines are pretty powerful, and when they want to go to a specific location, they’ll go there, and if your finger gets in between where it is and where it’s going, it’s going to be pretty dangerous.

How did the design of the robot evolve over time?

The robot design was a very collaborative process. It’s obviously a super-complicated robot, and so our design team came in to work closely with their mechanical engineers to understand the surgical process.

We started with the part of the robot that has the needle and is doing the actual insertion of the neural threads [which record brain activity], because it’s the most sensitive constraint, and we kind of worked backwards from there. We spent a period of time with them designing the part of the robot that interfaces with your head. We had to understand all the ways that you’d have to assemble it to cover up the existing system underneath it.

We then moved on to the rest of the robotic body, and we were able to develop the body in parallel with their internal electromechanical design team. We were able to order the units to be fabricated, and then we worked with them to assemble it. From there, they’ve taken it and done further internal testing.

What interests you about designing neurotech devices?

I’m always inspired by the people doing work in this space in terms of founders, scientists, technologists, neuroscientists, and personally it’s just really cool that the feat of this technology is unlocking big philosophical questions about how the brain works and what it means to be human. I think that’s super cool.

You’ve worked with other brain device companies. Are there particular use cases for neurotech that really excite you?

The field is focusing on the most vulnerable right now, which is inspiring. The immediate attention is on how to help people who need help the most, such as those who are paralyzed, and the problems that are being solved are very direct. I think seeing more work being done on these problems with AI, having the AI solve those very practical problems, is what I’m the most excited about right now.

,

Then there’s also this question around how AI is going to impact Neuralink and neurotech in general. I think these more, kind of aspirational applications—like performance-focused applications that allow people to do things faster—are great, but I think that there’s still more work to be done to understand that area.

I’m still not sure what the end user experience is going to look like when AI comes together with neurotech and robotics. What’s an optimistic future that is human-centered that we can think about? I’m all for humanity. I’m optimistic, but there’s also a part of me that’s a little bit concerned and has a fear of dystopian futures right now.

Tell me more about those fears.

Yeah, definitely. It’s gotten a little more concerning as we’re moving into conversations about AI and all the data necessary to make AI work. Data is gold, and real-time data is even more important. There are new consumer-facing [brain-computer interfaces] which are basically gathering data in real time to create better experiences. We’re slowly kind of walking into territory where we’re once again discussing giving up our privacy or data to companies. We’re on the verge of companies needing to know so much to do these new applications. It pushes against this idea of our own cognitive liberty. If you’ve got these AIs in a neurotech device that are helping you make decisions, what’s their best interest? Is it you? Is it the companies that built them?

Would you ever turn down a project if you felt like you didn’t agree ethically with what they were building?

Yeah, for sure. It’s a little bit tricky. In the practical world, there’s only so much that our clients share with us. We’re consultants at the end of the day, and so we don’t really have their long-term business roadmap at our disposal, as lovely as that would be. But I think understanding the background of the founder and their general worldview is useful. And I think that people who start companies are also very interesting people. They’re very ambitious, and sometimes you can’t differentiate between ambition and nefariousness.