CRUISE Sensor Pod: Design for the Reveal, Redesign for the Road (Noise)

CONCEPT ART, COMPONENT LAYOUT, 3D MODEL, RENDERINGS

DISCIPLINE(S): Industrial Design
ROLE: Principal Industrial Designer (sole designer; full-scope ownership from concept through production handoff)
SKILLS & TOOLS: Sketching, Photoshop, Class A Surfacing, Solidworks, Keyshot

Cruise, a San Francisco company developing a self-driving car system, needed a roof-mounted electronics enclosure to house a radar, two cameras, and their supporting components. I owned that enclosure through two different chapters, one where it had to stand on its own, and one where real-world testing forced it to change.

getcruise.com ⭧

Problem/Opportunity‍ ‍

The enclosure needed to house specific hardware. The brief was noncommittal on styling, no direction either way on whether the enclosure should reflect the host car's design language or stand apart from it.
Once the first version reached real-world road testing, a problem invisible at the concept stage surfaced: audible wind noise at highway speed, a functional failure no amount of aesthetic refinement could paper over.

‍ ‍

Goal‍ ‍

Version 1: design a distinctive, production-viable enclosure, with the styling question genuinely open, free to explore forms that stood apart from the host vehicle or echoed it.
Version 2: eliminate the wind noise problem through form change, while shifting the enclosure's aesthetic to blend with the specific test vehicle it rode on, a direct reversal of the original approach.

‍ ‍

Impact‍ ‍

The Version 1 enclosure was 3D-printed, finished, and painted, real hardware, and photographed, but those photos weren't widely released. Full public reveal came later, with Version 2.
Version 2 solved a measurable functional problem and completed a full identity reversal in the same set of revisions, from an independent design language to close visual integration with the vehicle underneath it, and it was this version the public actually saw.

Version One

An Independent Identity

The brief started as a packaging problem: fit a radar, two cameras, and their electronics into a single roof-mounted enclosure, with no mention either way of whether it should echo the host car's styling or read as its own thing. Sketches and photomanipulations explored a wide range of forms in that open space, treating the enclosure as its own object rather than defaulting to an accessory look.

*The concept eventually chosen to become Version 1 color-matched, rendered, and photoshopped onto an existing image.*

Initial Model

3D models followed from both the strongest sketches and fresh starts developed straight in CAD. The team converged on one direction, color-matched, rendered, and composited onto real vehicle photography to judge it in context before committing to a physical build.

The selected concept became a real, physical object, 3D-printed, finished, and painted, not a rendering. It was photographed, but those images stayed internal rather than going out publicly, the design still had real-world testing ahead of it before Cruise was ready to put it in front of an audience.

Version Two: What the Road Found

The Problem Concept Review Couldn't Catch (Rev 2A)

Real-world testing surfaced something no amount of design review had: at highway speed, the gap between the enclosure's side wings and the vehicle roof beneath them generated audible wind noise. This wasn’t a problem a styling pass would fix, and it required closing the gap and rethinking the underlying aerodynamic form.

Revision 2A addressed the noise directly by improving aerodynamics and closing that gap, and used the same pass to question whether the radar glass was even necessary to expose visually — cutting it from the design along with reduced overall width and length.

Matching the Vehicle (Rev 2B)

The final revision turned entirely toward aesthetics, refining the enclosure to closely match the styling of the Audi test vehicle it was mounted to, rather than reading as a distinct, independently-designed object bolted on top. The identity that had deliberately stood apart in Act One now deliberately blended in. This was the version that finally went public, featured in Cruise's reveal advertising once both the functional and aesthetic work were done.

*Promotional photos taken on reveal day*

The Throughline

V1 looked finished. It was photographed, painted, production-ready by every visible measure, and still wasn't shown to the public, real-world testing found the wind noise problem before Cruise ever did. That's the real story here, not the enclosure's final shape. Being camera-ready and being road-ready turned out to be two different bars, and only one of them showed up in a design review.

What's worth noting is what the fix turned out to be: closing the wind-noise gap was an aerodynamic correction, but it also simplified and cleaned up the form, meaning the functional problem and the aesthetic one weren't actually separate problems wearing different names. The identity flipped entirely too, from an object built to look like nothing else on the road, to one built to disappear into the vehicle carrying it. Only once both were resolved, function and integration together, was the enclosure actually ready for its debut.