Logistics / Autonomy · Deep dive

Nuro, Inc.

Mountain View autonomy company founded in 2016 by ex-Waymo engineers Dave Ferguson and Jiajun Zhu that raised $2.13B through 2021 at an $8.6B peak valuation, then re-cut itself twice — 20% in November 2022, 30% in May 2023 — abandoned its custom R2/R3 delivery pod, and pivoted in September 2024 to licensing the Nuro Driver Level-4 stack to OEMs and mobility platforms; the July 2025 Uber–Lucid–Nuro deal for 20,000+ Lucid Gravity robotaxis over six years is either the resurrection or the final chapter.

emerging

The question that decides it: Can autonomy-licensing revenue from the Uber–Lucid Gravity program scale fast enough between 2027 and the end of Nuro's cash runway to make the company a durable Tier-1 AV supplier — or does Nvidia's own DriveOS reference stack, Mobileye Chauffeur, and Waymo's first-party expansion pin Nuro as a single-customer integration partner that the OEMs and platforms eventually route around?

My take ★★ Very interesting ★ Interesting − Maybe ✗ Not interesting

Snapshot

Nuro is a Mountain View, CA autonomy company founded in June 2016 by Jiajun Zhu and Dave Ferguson, two of the earliest engineers on what is now Waymo. Between 2017 and 2021 it raised roughly $2.13B — including SoftBank Vision Fund’s $940M in February 2019 at $2.7B pre and a $600M Tiger Global / Google / Fidelity round in November 2021 at $8.6B — to build custom driverless delivery pods (R1, R2, and the never-mass-produced R3) that would deliver Kroger groceries, Domino’s pizzas, 7-Eleven convenience orders and FedEx parcels. By late 2022 and mid-2023 the model had broken. Nuro cut 20% of staff in November 2022, another 30% in May 2023, shelved R3 volume production, and in September 2024 formally pivoted from operating its own pods to licensing the Nuro Driver Level-4 autonomy stack to OEMs and mobility platforms. In April 2025 it raised a $106M Series E at a down-round $6B valuation, extended in August 2025 to $203M with Nvidia and Uber joining. The July 2025 six-year Uber–Lucid–Nuro deal for 20,000+ Lucid Gravity robotaxis — Houston commercial launch mid-2027, SF sooner — is the largest single announced commitment to Nuro’s licensing model. Whether it is a resurrection or a single-customer trap is the entire investment question.

Founding story

Nuro’s founding is a Waymo diaspora story. Jiajun Zhu joined Google’s self-driving project in 2009 as a founding-team principal software engineer, running perception and simulation and accumulating more than 100 patents (Nuro; Trium Group, 2024). Dave Ferguson joined the same project in 2011 as principal ML engineer, building and leading computer-vision, machine-learning, behavior-prediction and scene-understanding teams; he had earlier led planning for CMU’s Tartan Racing team that won the 2007 DARPA Urban Challenge. Both left in April 2016 amid the internal reorganisation that would produce Waymo as a separate Alphabet subsidiary the following December.

They incorporated Nuro in June 2016 with a specific, uncontroversial thesis: passenger robotaxis were regulatorily and technically further from deployment than goods-only delivery vehicles, because a delivery pod could be occupant-free by design, geo-fenced to slow suburban roads, and operate at up to 25 mph. Less risk, easier NHTSA framing, an obvious customer set (grocery, pharmacy, quick-service restaurants, parcel). Greylock led a $92M Series A in June 2017, followed by SoftBank Vision Fund’s $940M in February 2019 at $2.7B pre — the largest single check ever written into an AV-delivery-first company (Cooley; TechCrunch, 11 February 2019). In November 2019 NHTSA granted Nuro the first-ever federal exemption from Federal Motor Vehicle Safety Standards for a driverless vehicle, allowing the R2 pod to ship without side mirrors, windshield or occupant-facing crash structure. That regulatory win is still the most novel item in Nuro’s history and is a durable moat asset even after the strategic pivot.

How it works

The Nuro stack is a full Level-4 autonomy system: sensor fusion, HD-map plus real-time perception, prediction, motion planning, and a fleet-teleoperations backstop. In the original R2/R3 architecture the stack ran onboard a proprietary pod: LiDAR, radar, cameras, thermal, ultrasonics; a compute rack; and cargo compartments that a customer would unlock at kerbside via a code on their phone. R2’s operational design domain was low-speed suburban surface streets, initially in Houston and Phoenix, later Mountain View, in stable weather and mapped geofences.

Since the September 2024 pivot the same stack is being repackaged as Nuro Driver: the sensor suite, compute module (Nvidia DRIVE AGX Thor from 2025), autonomy software, and the fleet-ops toolchain are integrated into a partner OEM’s vehicle platform. The first announced integration is Lucid’s Gravity SUV — a passenger vehicle Nuro would not have built four years ago — with Uber as the dispatch layer. Nuro is publicly running ~100 test vehicles across the Bay Area and Houston (Nuro press, 2026) and got a California DMV Driverless Testing Permit for the Lucid Gravity in May 2026, joining only five companies with both that permit and a CPUC drivered-pilot permit (TechCrunch, 5 May 2026). NVIDIA GPUs in the cloud run the training and simulation stack; Uber runs dispatch and fleet operations; Lucid manufactures the vehicles. Nuro’s role is autonomy software plus safety case plus HD map ownership plus first-line teleoperations. This is a considerably narrower operational surface than the 2020-2022 “own the pod, own the deliveries” strategy.

Product and business overview

Nuro Driver. The Level-4 autonomy platform. Sensor kit, compute module on NVIDIA DRIVE AGX Thor, stack software, HD map, fleet dashboard, teleoperation. Sold to OEMs and mobility platforms under multi-year integration deals.

Nuro Delivery Bot (R2 / R3 — legacy). The custom driverless pod. R2 is what actually operated commercially 2020-2023; R3 was announced 2022 and shelved in the May 2023 restructuring. The company still ships and services R2 vehicles for select R&D and demo purposes but no longer positions the pod as its primary go-to-market.

Testing operations. ~100 driverless test vehicles operating 24/7 with and without safety operators across the Bay Area and Houston (Nuro press, 2026). The California DMV Driverless Testing Permit and the CPUC drivered-pilot permit are competitive scarcities.

Commercial partnerships (active). Lucid + Uber for the Gravity robotaxi program (July 2025, six-year, 20,000+ vehicles). No publicly disclosed second flagship licensee as of Q3 2026. Historical customers Kroger, Domino’s, 7-Eleven and FedEx are described in company materials but have not appeared in commercial-partner disclosures since 2023.

Business model and pricing

Pre-2024 the model was per-delivery fees from consumer-goods retailers using the R1/R2 pod: not disclosed in dollars but industry press reported per-order economics comparable to human delivery only at very high pod utilisation. Post-2024 the model is autonomy-stack licensing plus integration NRE plus per-mile / per-fleet-vehicle usage — the same three-part revenue structure Aurora Innovation uses for freight and Mobileye uses for OEM ADAS-to-L4.

Real prices are not published. The publicly available anchor is the Uber commitment: Uber invested “multi-hundred-million dollar” investments in Nuro in July 2025 alongside a $300M direct investment in Lucid, and Uber has said aggregate spending across the six-year, 20,000-vehicle Lucid Gravity + Nuro Driver program supports a Houston mid-2027 launch and dozens-of-market expansion after that (CNBC, 17 July 2025; Uber investor release, July 2025). No unit-economic disclosure of the per-vehicle license, per-mile royalty or gross-margin split among Nuro / Lucid / Uber has appeared in public filings.

Cash position: Nuro disclosed ~$1.3B on the balance sheet at the May 2023 restructuring announcement — enough for a three-year runway to mid-2026 (TechCrunch, 12 May 2023). The April and August 2025 Series E tranches added $203M gross, extending nominal runway to ~2028 on 2024-vintage burn assumptions. The pivot to licensing is designed to lower burn (no pod manufacture, no fleet operation cost centre) but ramps revenue only when partner vehicles start monetising, which for Lucid Gravity is 2027 at earliest.

Traction over time

Sources: TechCrunch (2019-2026); Bloomberg; NHTSA; CA DMV; company press releases; The Robot Report.

The pattern is a bell curve. Capital, headcount, permit surface and commercial-partner count peaked between 2021 and 2022, contracted 2022-2023, and are being rebuilt from a narrower base around a single licensing thesis in 2024-2026.

Market analysis

The two markets Nuro touches are very different in size and structure.

Autonomous last-mile delivery — Nuro’s 2016-2023 home turf — is estimated by Fortune Business Insights at ~$10.13B in 2026 growing to ~$212B by 2034 under aggressive scenarios; Global Market Insights forecasts a narrower cut at $1.6B in 2026 to $11.5B by 2035 at 24.5% CAGR (Fortune Business Insights; Global Market Insights, 2026). Real deployments remain concentrated in sidewalk-scale robots (Serve, Starship, Coco, Cartken) rather than street-scale pods. Nuro’s implicit view since 2024 is that the street-scale pod economics don’t work without vehicle volumes that require an OEM partner rather than a bespoke build.

Robotaxi / Level-4 passenger autonomy — where the July 2025 Uber–Lucid–Nuro deal points — is a much bigger prize with a much bigger set of competitors. Fortune Business Insights values the 2026 global robotaxi market at $1.27B growing to $96.31B by 2034 at 71.9% CAGR. Waymo has already scaled a first-party service in Phoenix, SF, LA, Austin and Miami; Zoox is ramping Las Vegas, SF and Austin; Tesla is building a first-party Robotaxi service on its own vehicle base. Nuro’s specific bet is that OEMs and platforms who don’t want to build their own stack will license one — and that Nuro Driver is the choice against Aurora (freight-focused), Wayve (earlier-stage), Mobileye Chauffeur (encumbered by OEM Tier-1 relationships) and Nvidia’s own reference stack.

The structural force to watch is Nvidia. Nvidia is Nuro’s compute supplier, an August 2025 cap-table investor, and — through DRIVE OS and the DRIVE Halos safety framework — an increasingly complete reference software stack that any OEM can adopt directly. In every other Nvidia software category the reference stack has eventually cannibalised third-party middleware. Whether autonomy is the exception is the central uncertainty.

Competitive intel

Waymo is the founders’ alma mater, the L4 leader in commercial passenger deployments, and — via a separate Uber commercial partnership in Phoenix and Austin — a direct alternative supplier to Uber for the same demand Nuro is selling into. Zoox is Amazon-owned and sits on the world’s largest e-commerce logistics stack; the R1/R2 delivery-pod thesis Nuro pioneered rationally belongs inside Zoox now. Aurora Innovation (AUR) is the direct comp for the licensing pivot — founder-and-CTO team also ex-Waymo (Chris Urmson, Sterling Anderson), publicly traded, further ahead on freight commercialisation with PACCAR and Volvo. Kodiak AI attacks the L4 licensing lane from military and long-haul freight. Wayve brings an end-to-end learning pitch that resonates with 2025-2026 AI-first buyers and is closer to Nvidia culturally. Mobileye (MBLY) is the incumbent OEM ADAS-to-L4 supplier with existing contracts inside every OEM Nuro would sell to. Applied Intuition sells the tools stack Nuro would displace if OEMs bought whole stacks; if OEMs prefer tools, Applied wins. Serve Robotics (SERV), Starship, Coco, Cartken and Avride together own the sidewalk-scale delivery-robot lane Nuro chose not to defend.

The most dangerous competitor is the least visible: Nvidia DRIVE. Nvidia’s compute is the substrate for Nuro Driver and for many of the OEM programs Nuro would license to. Nvidia’s own reference software has expanded release-by-release. Nvidia is also an equity holder in Nuro — which cuts both ways: strategic customer today, potential displacer tomorrow.

History and evolution

• 2007 — Ferguson leads planning for CMU Tartan Racing DARPA Urban Challenge win.
• 2009 — Zhu joins Google’s self-driving car project as founding-team principal software engineer.
• 2011 — Ferguson joins Google’s self-driving project as principal ML engineer.
• April 2016 — Zhu and Ferguson leave Waymo.
• June 2016 — Nuro incorporated in Mountain View.
• June 2017 — $92M Series A led by Greylock and Banyan.
• June 2018 — Kroger partnership announced.
• December 2018 — First R1 grocery deliveries in Scottsdale, AZ.
• 11 February 2019 — $940M Series B led by SoftBank Vision Fund at $2.7B pre.
• November 2019 — NHTSA grants Nuro first-ever driverless-vehicle FMVSS exemption for R2 (final rule February 2020).
• April 2020 — R2 pod unveiled; CVS Houston prescription deliveries begin.
• 9 November 2020 — $500M Series C led by T. Rowe Price at $5B.
• 12 April 2021 — Domino’s Houston autonomous pizza pilot launches.
• 2 November 2021 — $600M Series D led by Tiger Global, Google joins, at $8.6B (peak valuation).
• 1 December 2021 — 7-Eleven California autonomous delivery pilot launches.
• 20 January 2022 — FedEx Houston multi-phase last-mile pilot begins.
• 14 July 2022 — Winds down Phoenix operations.
• 18 November 2022 — First layoff: ~300 heads (20%).
• 12 May 2023 — Second layoff: ~340 heads (30%); R3 volume production delayed indefinitely; commercial expansion paused.
• 11 September 2024 — Strategic pivot: autonomy stack to be licensed rather than deployed in first-party pods.
• 19 November 2024 — Nuro expands driverless testing across Bay Area and Houston.
• 9 April 2025 — Series E first tranche: $106M at $6B (down-round from $8.6B).
• 17 July 2025 — Uber–Lucid–Nuro six-year, 20,000+-vehicle robotaxi program announced; Uber invests multi-hundred-million into Nuro and $300M into Lucid.
• 21 August 2025 — Series E second tranche: $97M with Nvidia and Uber as new investors; total Series E $203M at $6B.
• January 2026 — Global-robotaxi unveil at CES 2026; first prototype operating autonomously at Nuro’s Las Vegas proving ground.
• 13 April 2026 — Uber begins early Bay Area test rides of Lucid Gravity + Nuro robotaxi.
• 5 May 2026 — California DMV grants Nuro Driverless Testing Permit for Lucid Gravity.
• July 2026 — Houston 2027 launch confirmed; 50,000 sq ft Uber Houston depot secured.

What people say

The case for. Nuro is the surviving second-generation Waymo diaspora company with a full L4 stack, first-ever NHTSA driverless FMVSS exemption, real driverless testing miles in Bay Area and Houston, one of only five companies holding both a CA DMV Driverless Testing Permit and a CPUC Drivered Pilot Permit (Nuro; TechCrunch, 5 May 2026), Nvidia and Uber on the cap table, and a 20,000-vehicle six-year commercial anchor (Uber IR, July 2025). Glassdoor shows an overall 4.2 / 5 rating across ~370 reviews with 83% recommending the company, 4.3 for culture and values and 4.1 for career opportunities (Glassdoor, 2025). Employees regularly cite the technical talent bar and cross-functional collaboration as reasons to stay. The Contrary Research memo argued in 2023 that Nuro’s willingness to shed the pod-manufacturing millstone was the right strategic call and the survivable path (Contrary Research, 2023).

The complaints. Two mass layoffs in six months (November 2022; May 2023) totalling roughly half the workforce; the flagship R3 pod product shelved indefinitely; a 30% down-round in April 2025 from the 2021 peak of $8.6B to $6B — a haircut that reprices vested employee equity at 70 cents on the dollar even before the ordinary preference stack (TechCrunch, 9 April 2025). Glassdoor 2024-2025 reviews describe “leadership constantly changing their plan,” “executive leadership indecision to pivot company direction to follow AI trend words,” and an 80% RTO policy announced after business hours without prior consultation (Glassdoor, 2024-2025). The commercial-partner list Nuro highlighted at peak — Kroger, Domino’s, 7-Eleven, FedEx — has not surfaced in any 2024-2026 disclosure as continuing at scale; the July 2022 Phoenix wind-down effectively ended the Kroger relationship, and industry press has not reported active Domino’s or 7-Eleven deployments since 2023 (TechCrunch, 14 July 2022; Nuro press archive). Ride AI and TechCrunch coverage of the licensing pivot noted an “existential” quality to the shift, with Uber-and-Lucid effectively the single anchor demand-side commitment. The single-customer concentration risk is now the loudest quiet complaint: if the Lucid Gravity ramp slips, the 2027 launch dates slip, or Uber routes to Waymo or Zoox instead, there is no visible second flagship.

Outlook: the open question

The answer condition is time versus displacement. Nuro works if — and only if — the July 2025 Uber–Lucid Gravity program (i) actually delivers commercial revenue in Houston by mid-2027 and San Francisco in the same window, (ii) hits enough deployed vehicle count between 2027 and 2029 to convert the licensing model from R&D-heavy to a recurring per-mile / per-vehicle royalty stream, and (iii) attracts at least one credible second OEM licensee — not another retailer partnership but a second, non-Lucid vehicle platform running Nuro Driver — before the end of the current $2.33B capital-raised runway (implied late-2028 / early-2029 on 2024-vintage burn). If those three conditions are met the company is a viable Tier-1 AV supplier at a $10-20B outcome. If any of the three fails, the more likely end state is an acquihire into Nvidia, Uber or a Tier-1 supplier at a valuation at or below the 2025 $6B post — the kind of quiet unwind Cruise and Motional have already priced for the sector.

What flips the call the wrong way. (1) Nvidia’s own DRIVE reference stack becomes a full L4 offer OEMs adopt directly — Nuro’s compute partner cannibalises the software layer above. (2) Waymo, whose stack was built by the same team, keeps expanding first-party in the exact cities Uber is opening for Nuro — Houston, SF, Austin — and Uber routes through the Waymo commercial relationship instead. (3) The Lucid Gravity ramp encounters normal automotive delays and the 2027 Houston launch slips to 2028 while burn continues. (4) The safety case takes a public setback of any kind — a Cruise-type October-2023 incident would break the entire pivot before it earns a dollar.

What flips it the right way. A second signed OEM licensee announced by end of 2027; disclosed per-vehicle Nuro Driver economics in an S-1 or private disclosure that show ≥30% gross margin at scale; and public rides in two US cities generating >100k paid trips a month by end of 2028. Those are the three quantitative gates.

How to attack it

The specific wedge is not another L4 stack for OEMs. That fight is Aurora versus Wayve versus Mobileye versus Nvidia, and it costs ~$500M per contender to be credible. The attackable wedge is the lane Nuro left: street-scale, purpose-built, single-tenant delivery pods for high-density QSR, grocery, pharmacy and hyper-local logistics customers, deployed on an operator model rather than a licensing model. Domino’s, 7-Eleven, CVS, Kroger, Amazon Fresh, Walmart, Chick-fil-A and every dark-store operator still want lower per-delivery cost at 2-5 mile radius; sidewalk robots are too slow and too small; human delivery is too expensive; Nuro conceded the street-scale-pod thesis in 2024. A well-funded new entrant could re-open exactly that lane using Nvidia DRIVE compute (commoditised since 2025), OEM-built commercial vehicle platforms (Ford Transit Connect, Rivian EDV variants) rather than bespoke manufacture, off-the-shelf lidar (Ouster, Innoviz) at 2026 prices roughly 60% below 2020, and an operator business model billing per delivery — precisely the P&L Nuro walked away from.

Exploitable weaknesses. (i) Single-customer commercial concentration: Uber–Lucid is effectively the entire disclosed anchor (CNBC, 17 July 2025). (ii) Down-round overhang from the April 2025 $6B mark rerating from the $8.6B 2021 peak — makes hiring senior technical talent against fresh-mint AI-startup RSU packages harder (TechCrunch, 9 April 2025). (iii) NVIDIA is both supplier and cap-table investor and increasingly the direct competitor — a structural conflict that will surface the moment Nvidia sells DriveOS L4 to a Nuro OEM prospect. (iv) Public partnerships list has decayed from Kroger / Domino’s / 7-Eleven / FedEx / CVS in 2021-2022 to Uber–Lucid in 2025-2026; the sector reads that as churn. (v) 80% RTO enforcement in the Bay Area with a “3-days-in-office” policy criticised in 2024-2025 Glassdoor reviews as top-down (Glassdoor, 2024-2025). (vi) The NHTSA FMVSS exemption is R2-specific and does not immediately transfer to Lucid Gravity, which requires a separate FMVSS compliance pathway as a standard passenger vehicle. (vii) No disclosed second OEM licensee 14 months after the September 2024 pivot. (viii) Contrary Research and Ride AI have both explicitly framed the pivot as “existential”, meaning the market prices Nuro’s optionality below its remaining cash (Contrary Research, 2023; Ride AI, 2025). (ix) 2022-2023 layoffs still show up as institutional knowledge loss inside every Glassdoor comment about product-management indecision.

Adjacent-segment play

The most attractive adjacent segment is AV data and simulation infrastructure. Nuro has spent nine years and roughly $2B accumulating a proprietary dataset of driverless miles, edge-case interventions, HD-map artifacts and synthetic-scenario libraries — the exact input every OEM autonomy program and every AI foundation-model builder now wants for embodied-AI training. Applied Intuition ($6B, 2024) sells the tooling; Foretellix, Parallel Domain and Cognata sell simulation. A Nuro spin-out or partnership repackaging its dataset and scenario library as a subscription product to the top 15 OEM autonomy programs plus every physical-AI foundation-model shop (Physical Intelligence, Skild AI, 1X, Figure) would generate revenue that is orthogonal to whether the Lucid Gravity program hits its 2027 dates.

A second adjacent worth serious consideration is embodied-AI beyond ground vehicles: warehouse robots, humanoid logistics, delivery drones. Nuro’s stack combines the geometric-perception primitives, prediction models and motion-planning software that transfer directly to physical-intelligence startups. A specific spin-out for warehouse or micro-fulfillment autonomy could pick off a piece of the $30B+ warehouse-automation TAM (Symbotic, GreyOrange, AutoStore reference cases) using an accumulated advantage no purely software-first entrant has.

A third adjacent — narrower — is defence and dual-use ground autonomy. Anduril, Shield AI and Kodiak have shown that AV stacks generalise to unmanned ground vehicles under DoD budgets. Nuro’s Palo Alto proximity, permit surface and CMU/Waymo pedigree slot into a dual-use pitch that would separately monetise a stack whose commercial revenue is speculative to 2028. If Uber and Lucid disappoint, this is the natural pivot Nuro’s next investors would pay for.

Sources and further reading

• Nuro Closes $203 Million Series E Financing to Advance Its AI-First Self-Driving Technology and Commercial Partnerships — BusinessWire, 21 August 2025.
• Nuro’s $106M raise backs its shift from delivery robots to licensing autonomy tech — TechCrunch, 9 April 2025.
• Uber inks six-year robotaxi deal with Lucid, invests $300 million in EV company — CNBC, 17 July 2025.
• Lucid, Nuro, and Uber Partner on Next-Generation Autonomous Robotaxi Program — Uber investor release, 17 July 2025.
• Autonomous delivery startup Nuro to lay off 30% of workforce — TechCrunch, 12 May 2023.
• Autonomous vehicles startup Nuro winds down operations in Phoenix — TechCrunch, 14 July 2022.
• Nuro pivots to license self-driving tech to carmakers, mobility companies — TechCrunch, 11 September 2024.
• Nuro lands investment from Google, Tiger Global in new $600M round — TechCrunch, 2 November 2021.
• SoftBank Vision Fund Invests $940 Million in Nuro — Cooley (deal counsel), 11 February 2019.
• Nuro receives driverless testing permit ahead of Uber robotaxi service launch — TechCrunch, 5 May 2026.
• Uber, Nuro, and Lucid to Bring Robotaxi Service to Houston in 2027 — Uber investor release, July 2026.
• Nuro — Contrary Research — Contrary Research memo, 2023.
• Nuro Reviews — Glassdoor — Glassdoor, 2024-2026.
• Nuro Closes $203M Funding Round, Nvidia and Uber On Board — Ride AI, August 2025.
• Autonomous Last Mile Delivery Market Size Report, 2034 — Fortune Business Insights, 2026.
• Robotaxi Market Size, Share — Industry Report 2026-2034 — Fortune Business Insights, 2026.