Here's a number that should reframe how you think about the autonomous vehicle race: the sensors that make self-driving viable cost somewhere in the "low hundreds of dollars" — and virtually every competitive option at that price point is manufactured in China. That's not a complaint from a mid-tier supplier. That's Rivian CEO RJ Scaringe, speaking to Reuters on Tuesday, explaining why his company is now seriously weighing whether to manufacture its own lidar sensors on American soil.

Rather than buy directly from a Chinese supplier, Rivian is mulling making its own lidar sensors in the United States using Chinese technology, possibly through a joint venture, Scaringe said. Chinese suppliers such as Hesai Group and RoboSense have come to dominate the market for smaller, cheaper lidar sensors, but the rise of Chinese sensors has raised national security concerns among U.S. lawmakers.

That's the buried lede. Rivian (RIVN) mulls making its own lidar not primarily because it's a capability play — though it is — but because the geopolitical ground beneath the entire autonomous vehicle supply chain is shifting fast, and every automaker building serious self-driving ambitions is caught in the same bind.

The China Chokepoint Nobody Talks About Openly

Scaringe's Reuters interview is notable precisely because he said what most automakers are only thinking. "All the real choices are coming out of China" for the low hundreds of dollars price point that automakers require, Scaringe said. That's a blunt acknowledgment of a structural dependency that U.S. legislation hasn't fully caught up to — but is trying to.

The two companies at the center of this dynamic are worth understanding. RoboSense's average ADAS selling price fell to approximately 1,500 yuan — about $218 — per unit in Q4 FY2025. The leading automotive lidar companies include RoboSense, Hesai Technology, Luminar Technologies, Innoviz Technologies, Cepton (now owned by Koito), and Aeva Technologies. Luminar supplies Volvo and Mercedes-Benz. The American players in that list — Luminar, Innoviz, Aeva — are not operating at the price points or production volumes of their Chinese counterparts. That gap isn't closing quickly.

RoboSense said in April 2025 it had partnered with "a leading emerging automaker in North America" and a "North American new energy vehicle brand." The timing aligned almost exactly with Rivian's December 2025 announcement that it would integrate lidar into the R2. Rivian officially denied RoboSense is the supplier, but the circumstantial evidence pointed clearly enough that the industry took notice. Now, with Scaringe publicly discussing domestic manufacturing through a potential JV with a Chinese firm, the subtext has become text.

Scaringe said Rivian is in "active discussions" with lidar firms and that the effort might also include other automakers. "A number of different car manufacturers are thinking about how they could do that either together, or at least through a shared alignment to say, hey, let's develop production capacity in the United States for this, or at least outside China," Scaringe said.

A multi-automaker lidar manufacturing consortium, using Chinese technology but built in the U.S., is a genuinely novel structure — and if it materializes, it would represent the first major collaborative hardware response to the China dependency problem in autonomous driving.

What Rivian Has Already Built

The lidar announcement doesn't land in isolation. It's the latest layer of a vertical integration strategy that Rivian has been executing quietly and, given the scale of its losses, expensively.

At its inaugural AI & Autonomy Day in December 2025, Rivian unveiled the RAP1 — a custom 5nm processor delivering 1,600 trillion operations per second of AI compute. The chip uses Arm's v9 architecture with 14 high-performance cores and is 2.5x more power-efficient than Rivian's previous systems. Rivian's Gen 3 Autonomy platform packs 11 cameras, 65 megapixels total, five radars, and one lidar sensor.

Senior vice president of electrical hardware Vidya Rajagopalan said at the event: "We expect that at launch in late 2026 this will be the most powerful combination of sensors and inference compute in consumer vehicles in North America."

The software layer is equally ambitious. The Rivian Autonomy Platform's Large Driving Model is trained end-to-end through reinforcement learning. Every LiDAR-equipped R2 sold will serve as a "ground-truthing vehicle," providing continuous training data for the company's self-driving system. Tesla's flywheel model — every vehicle a data collector — is the obvious reference point, and Rivian is explicitly replicating it.

The R&D spend tells the story in hard numbers. R&D expense was $458 million in Q1 2026 compared to $381 million in the same quarter the previous year, with the increase primarily related to increased software and cloud spend on autonomy and R2 pre-production costs. Annualize that and you're looking at close to $1.8 billion in R&D per year — from a company that just started producing its mass-market vehicle.

"What Rivian is attempting is genuinely rare among American automakers: true hardware-software co-design for autonomy. Most OEMs buy a stack and integrate it. Rivian is building the stack. The lidar question is the natural endpoint of that logic — if you're designing your own chip and training your own model, why would you accept dependency on a sensor you can't control? The answer is cost and timeline. Building lidar at scale is a decade-long project. Rivian doesn't have a decade."

— A senior autonomy researcher at a major U.S. university who consults for automotive OEMs and asked not to be named

The Uber Deal Raises the Stakes Considerably

None of this would carry the same weight without the commercial contract that now hangs over everything Rivian does on autonomy. Uber Technologies plans to invest up to $1.25 billion in Rivian as part of a deal to deploy up to 50,000 robotaxis across 25 cities in the U.S., Canada, and Europe through 2031. The tie-up includes expectations for Uber, or its fleet partners, to purchase 10,000 autonomous versions of Rivian's upcoming R2 electric vehicle, with the option to buy up to 40,000 more robotaxis beginning in 2030.

Scaringe said at SXSW 2026: "Our path to get to hands-off, eyes-off in 2027 is something we're spending more money on than anything else." That quote, combined with the Uber milestone structure — Uber's investment tranches are tied to Rivian hitting specific autonomous performance targets — means lidar reliability isn't an engineering preference. It's a cash flow event. If Rivian's sensor stack underperforms, the Uber capital doesn't arrive on schedule.

The company no longer expects to reach positive adjusted EBITDA by 2027, highlighting the cost-intensive nature of its autonomy push and platform expansion. That's a material concession. Rivian traded near-term profitability for the Uber deal and the full-stack autonomy build-out. The lidar manufacturing question is downstream of that choice.

The European dimension matters here too. Deploying robotaxis across 25 cities that include European markets means navigating the EU's AI Act and transportation safety frameworks — regulatory environments where sensor provenance and supply chain auditability carry legal weight, not just political optics. A domestically manufactured lidar sensor, even if the underlying technology is Chinese, positions Rivian more defensibly in those markets than a direct Chinese import would.

The skeptic's view deserves a paragraph. Rivian is simultaneously launching a new vehicle, building autonomous driving from scratch, expanding its Georgia plant to 300,000 units of annual capacity, managing a Volkswagen joint venture, and now apparently considering building its own sensor manufacturing operation. That's a capital allocation agenda that would strain a company with twice Rivian's liquidity. Its Q1 2026 free cash flow was negative $1.075 billion. Its total liquidity sits at $5.4 billion. These numbers are compatible with current plans — barely. Adding lidar manufacturing to the list, even through a JV, adds timeline risk and capital complexity at exactly the moment execution against existing commitments demands the most focus.

Who Wins, Who Loses

For Hesai and RoboSense, the signal from Scaringe is simultaneously a threat and a validation. Their technology is good enough that the CEO of a well-funded American EV maker is discussing using it as the basis for domestic manufacturing. Their competitive position in low-cost lidar is being directly acknowledged as unmatched. But the path to long-term contracts with U.S. automakers just got narrower, assuming Washington continues its current trajectory on connected vehicle restrictions.

For American lidar players like Luminar Technologies — which supplies Volvo and has struggled commercially while its Chinese competitors captured market share — a legislative or procurement mandate for domestically manufactured sensors would be a genuine lifeline. Rivian floating the concept of a multi-automaker consortium around domestic lidar production is, for Luminar, the most encouraging thing it's heard from a major OEM in years.

For Waymo and Tesla, the competitive read is nuanced. Waymo doesn't sell consumer vehicles; its sensor strategy and supply chain are insulated from the constraints Scaringe is describing. Tesla's camera-only approach, which Elon Musk has defended at length while dismissing lidar as unnecessary, sidesteps the problem entirely — though the autonomous performance debate between camera-only and sensor-fusion approaches remains genuinely unsettled. Adding in-house sensor manufacturing would give Rivian control over virtually every hardware and software component of its autonomy system. Tesla has that. Waymo has that. Almost no one else does.

Key Takeaways

The geopolitics are the real story. Rivian's lidar deliberations aren't primarily about capability — they're about supply chain resilience in a sector where U.S. lawmakers are increasingly scrutinizing Chinese automotive technology. Every automaker with serious autonomous ambitions faces this exact tension.

The Uber deal changed the calculus. Uber's investment is tied to the achievement of certain autonomous performance milestones — which means lidar reliability is now directly connected to Rivian's capital schedule. This is no longer a product decision. It's a financial one.

Vertical integration is Rivian's explicit strategy — and its biggest risk. Custom silicon, proprietary AI, a Large Driving Model, and now potential in-house lidar manufacturing. The coherence of this strategy is impressive. The capital requirements are terrifying, particularly with Q1 2026 free cash flow at negative $1.075 billion.

A multi-automaker lidar consortium would be genuinely significant. If Scaringe's public comments catalyze collective action among U.S. automakers to build domestic sensor manufacturing capacity, it would mark the first major industry-wide response to Chinese supply chain dominance in autonomous driving hardware.

The R2 launch is the hinge point. Rivian is guiding 62,000 to 67,000 deliveries for 2026. R2 saleable production just started. Until that vehicle reaches volume, every strategic ambition — robotaxis, lidar manufacturing, the Uber partnership — remains conditional on operational execution at a scale the company has never achieved.

Rivian (RIVN) mulls making its own lidar at the exact moment the company needs to be ruthlessly focused on producing cars, scaling autonomy, and managing cash. That it's even raising this question publicly — and apparently drawing other automakers into preliminary alignment — tells you something important: RJ Scaringe believes the sensor dependency problem is real enough that addressing it now, while painful, beats addressing it later when the geopolitical stakes are higher and the supply alternatives are fewer. Whether the capital and the timeline support that judgment is a question Rivian's next three quarters will start to answer.