Ouster: The Eyes of Physical AI

Every autonomous machine needs to sense its environment, and Ouster sells that capability rather than betting on any single end-market. Its sensors go into robotaxis, warehouse robots, drones, mining and farm equipment, traffic systems and defence platforms. This spread matters: no single application has to succeed for the business to grow, and several are scaling at once. As the wider industry concluded, safety-critical autonomy relies on layering cameras, radar and LiDAR together for redundancy, so rising autonomy spend pulls LiDAR demand up with it. Ouster is positioned to supply the perception layer across the whole field rather than one corner of it.

Ouster's Rev8 line, launched in 2026 with Fujifilm's colour expertise built in at the silicon level, is described as the world's first native-colour LiDAR, fusing depth and colour in a single sensor and removing the costly job of stitching camera and LiDAR data together. Crucially, Rev8 ships pre-integrated with NVIDIA's Jetson and DRIVE platforms, the standard tools developers already use to build robots and autonomous vehicles. That integration creates real switching costs: a developer building on NVIDIA finds Ouster the path of least resistance. Names reported to be adopting it include Google, Volvo Autonomous Solutions, Gecko Robotics and Skydio.

After years of losses, Ouster has shown it can grow revenue fast while improving margins, and it carries no debt with a healthy cash cushion. Two developments raise the quality of the business. A new defence channel opened when its OS1 sensor became the first high-resolution 3D LiDAR approved under the US "Blue UAS" framework, steering federally funded drone and infrastructure work toward Ouster while Chinese rivals are locked out by law. And the acquisition of Stereolabs, a 3D-camera and spatial-AI specialist, pushes Ouster from selling components toward owning more of the sensing-to-understanding software stack, which over time could lift margins and deepen customer lock-in.

• Rev8 design wins turning into orders: Conversion of named adopters like Gecko Robotics and Volvo Autonomous Solutions into volume shipments would reset the revenue trajectory above today's run-rate.

• Continued margin and growth proof points: Further quarters of double-digit revenue growth with stable margins would build confidence that profitability is durable, not a royalty-driven blip.

• Defence and infrastructure scaling: Blue UAS approval opens a federal drone and smart-city channel that barely existed a year ago, with US rules favouring domestic suppliers.

• Stereolabs and the software pivot: Integrating 3D-camera and perception software could add higher-margin, stickier revenue and move Ouster up the value chain.

• The physical-AI adoption wave: Broad rollout of robots, autonomous vehicles and smart infrastructure could lift LiDAR demand structurally, with Ouster positioned as a default perception layer.

• Capital discipline earning a re-rating: If management curbs dilution and compounds profitable growth, the market may award the business a higher-quality multiple.

Two forces squeeze the case from opposite directions. Lower-cost Chinese LiDAR makers such as Hesai and RoboSense compete hard on price, even if US procurement rules limit them in federal work. Meanwhile Tesla's bet that cameras and software alone can deliver self-driving remains a live philosophical challenge: if vision-only autonomy succeeds at scale, the argument that safe autonomy requires LiDAR weakens, and Ouster's premium with it. Most experts still favour sensor redundancy, but the debate is unresolved.

Ouster has crossed into profit only intermittently, and the numbers can mislead. A standout fourth quarter of 2025 (around 60% gross margin) was flattered by roughly $21m of largely one-time IP royalties; strip those out and the picture is more modest. In the first quarter of 2026, GAAP gross margin fell back to 43%, the company posted a net loss of $17.5m, and management guides to under $5m of royalties for the full year. Sustained, royalty-free profitability is still being built, and the path runs through holding margins as volumes grow.

At a roughly $2bn market value on annualised revenue still well under $250m, Ouster is priced for strong growth, not survival. That makes execution the whole game: the Rev8 design wins must convert from pilots into volume shipments, margins must hold, and the company must avoid heavy dilution. Ouster has repeatedly used its share-issuance facility when the stock rises, which can quietly expand the share count just as the story starts working, a pattern that frustrates investors who want stability.