Why the Neckband Is the Most Practical Bridge Toward True XR/AI Glasses

When people first imagine the future of XR or AI glasses, they often think of something lightweight and elegant — a pair of glasses as capable as a computer, yet as comfortable as wearing air. That vision is beautiful, but it’s still a few steps ahead of what our current technology can realistically achieve.
For now, the best bridge between today’s reality and that dream is something most people overlook — the neckband.

First of All, What Is a Neckband?

The neckband is a compact computing unit that rests comfortably around your neck. It powers your XR or AI glasses through a short wired connection, offering both the performance and stability needed for immersive experiences. VITURE, where I spent years working on XR products, has demonstrated how elegant and powerful this format can be. But the neckband isn’t the final destination — it’s a meaningful trade-off between comfort, power, and design.

© VITURE Inc.

Why Is the Neckband Better?

The biggest challenge in XR design is always weight. You can’t simply put a full computing system, large batteries, and active cooling inside a pair of glasses without turning them into a one-kilogram brick pressing down on your nose. Your head will protest long before the battery runs out. By contrast, the neckband carries most of that load comfortably at around 300 grams. The neck can bear weight much better than the face, and the distribution makes long sessions possible without discomfort. It also opens up room for beautiful, lightweight glasses that actually look like something you’d want to wear in public — not a prototype or a lab device.

There’s also an aesthetic argument. Connecting glasses to a TV box or a dangling cable hub is never elegant. The neckband, when designed well, feels integrated, intentional — a fashion accessory rather than a tech burden. It provides not only battery life and computing power but also the heart of the operating system. And in XR, a wired connection still matters. Display latency, stability, and visual fidelity remain significantly better through a cable than through wireless streaming, especially when rendering stereoscopic or high-frame-rate content.

From the software side, the neckband unlocks an entirely new layer of potential. It’s not just a “battery pack” — it’s an Android-based OS, meaning users can access millions of existing apps while developers can build new experiences tailored for XR. Even though traditional mobile apps aren’t fully optimized for spatial interaction — no hovering states, awkward D-pad navigation, and clumsy QR-code sign-ins — it’s still a thriving ecosystem to start from. Having an operating system allows us to dream bigger: to design our own apps, to build a developer community, even to run our own app store or AI voice agent. This is the kind of openness XR needs.

We began building partnerships with many developers years ago, including the creators of these remarkable streaming apps.

The neckband can also act as a motion sensor. In theory, combining its IMU data with that of the glasses helps detect subtle head or neck movements for more accurate spatial tracking. In practice, though, neck movement generates a lot of noise — it’s a challenge we faced during development and didn’t fully solve in my previous projects. But I still see great potential in this idea.

One More Thing to Share…

One of the most memorable stories from my time working on the neckband came from a frequent flyer. He told me he always wore our neckband on flights — not for XR viewing, but because he found it too heavy to pack separately. He never used the glasses during flights, but the neckband itself was always with him.

The folded Neckband. © VITURE Inc.

That conversation sparked an idea I never got to implement: a display-free mode, where you could still use the neckband for listening to music, podcasts, or real-time translations through voice interaction. With the built-in speaker and microphone — and support for wired or wireless headphones — the neckband could become a personal AI companion, even without the glasses. I dreamed about creating glasses and neckband working independently, yet better together. It’s a pity that I didn’t get to realize that idea.

What Did I Fail on the Neckband?

Of course, there are things I would do differently. We over-complicated interaction design. Gesture control was ambitious but unstable; the camera’s position on the neckband wasn’t ideal, and a single RGB sensor couldn’t capture reliable hand movements. In hindsight, the best place for gesture recognition is clearly on the glasses themselves. And we finally delivered a better and reliable gesture control system in 2025.

Gesture controls are intuitive. Image courtesy of A&C Systems’ LinkedIn post.

We also failed to provide a basic built-in control surface, like a small trackpad. Our most reliable control was the remote app, but that tied up users’ phones and sometimes felt frustrating. In user interaction, even a 1% error rate can ruin the magic.

Finally, we underestimated how fast AI-powered features would advance — our computing benchmark was conservative, and as a result, the neckband struggled with the performance demands of our AI 2D-to-3D model and SLAM algorithms. When both were running, heat and fan noise became unavoidable.

There Should Be an End to This Article

Despite these imperfections, I still believe the neckband is the most practical step forward toward the future of spatial computing. It gives us balance — between form and function, between wearability and power, between imagination and engineering reality. It may not be the destination, but it’s the bridge that will carry us there.

Clarify

This article was refined with the help of AI. Every opinion and story is my own.