Bay Area Robotics Map: Deep Dive on 46 Companies

The big picture first: This is not a random list. It is a snapshot of the "physical AI" capital bubble at its peak. Broken down from first principles, three things jump out.

• Money has detached from revenue. Figure at $39B, Skild at $14B, Mind Robotics at $3.4B in under a year, Sunday at $1.15B with zero robots in a single home. These are priced on founder pedigree and the hope that robot foundation models are the next large language model moment, not on shipped product. The single biggest recurring risk across the whole map is teleoperation dressed up as autonomy in demos.

• There is a civil war over where the value sits. The "brain" camp (hardware agnostic model labs like Physical Intelligence, Skild, Genesis, Dyna) versus the "body" camp (vertically integrated hardware makers like Figure, 1X, Tesla that build their own brains with proprietary data). Both sides bet the other is a commodity. They cannot both be right.

1. Robotaxi and autonomous vehicles

Waymo

• What: Fully driverless ride hailing using sensor rich retrofitted vehicles (the "Waymo Driver").
• HQ: Mountain View, CA. Alphabet subsidiary.
• Founded: Began 2009 as Google's self driving project, rebranded Waymo 2016. Co CEOs Tekedra Mawakana and Dmitri Dolgov.
• Funding: $16B round in February 2026 at a $126B valuation (Dragoneer, DST Global, Sequoia); Alphabet stays majority owner.
• Traction: 400,000+ paid rides per week across 6 US metros; roughly 15M rides in 2025; 20M+ lifetime. Expanding to 20+ new cities plus first international markets, London and Tokyo.
• Differentiator: The most mature, highest mileage driverless stack in commercial operation, redundant multi sensor suite (lidar plus radar plus cameras).
• Critical take: Clear leader, but per ride profitability is still unproven at a $126B mark. Tesla's camera only approach threatens on cost if it ever reaches safety parity. Is this a durable margin business or a subsidized land grab?

Zoox

• What: Purpose built bidirectional robotaxi with no steering wheel or pedals, run as an Amazon owned service.
• HQ: Foster City, CA.
• Founded: 2014 by Tim Kentley Klay and Jesse Levinson. Acquired by Amazon in 2020 for ~$1.2B; now funded internally.
• Traction: Public rides launched on the Las Vegas Strip (September 2025) and in San Francisco; Uber distribution partnership in Las Vegas; building a Hayward factory targeting up to ~10,000 vehicles/year. Still pre revenue on paid fares as of late 2025.
• Differentiator: Ground up robotaxi, no traditional driving controls, symmetric bidirectional driving, integrated sensor and compute stack.
• Critical take: Years behind Waymo on commercial scale and still pre revenue on paid rides. The purpose built vehicle is capital intensive and needs novel NHTSA approval for a car with no conventional controls. Amazon's balance sheet is the main reason to believe.

2. Autonomous trucking (middle mile)

Gatik

• What: Level 4 autonomous box trucks on short fixed routes between distribution centers and stores.
• HQ: Mountain View, CA.
• Founded: 2017 by Gautam Narang, Arjun Narang, Apeksha Kumavat.
• Funding: ~$270M total (Koch Industries among backers); valued above $800M.
• Traction: Walmart, Tyson, Kroger, PepsiCo (~250 retail locations). Fully driverless since mid 2025; reports ~$600M of contracts over five years.
• Differentiator: Deliberately narrow operating domain (short, fixed, repeatable routes) makes full driver out autonomy bankable now.
• Critical take: Smart for near term revenue but caps the total market, and it leans on a handful of large retail and consumer packaged goods accounts (concentration risk). Can it defend middle mile margins as long haul players mature?

PlusAI

• What: "SuperDrive" Level 4 virtual driver software sold to truck makers rather than run as a fleet.
• HQ: Santa Clara, CA.
• Founded: 2016 by David Liu, Shawn Kerrigan, Hao Zheng.
• Funding: Going public via SPAC with Churchill Capital Corp IX, expected to close around February 2026, ~$1.2B valuation and ~$300M proceeds.
• Traction: OEM partnerships are the core: TRATON Group (Scania, MAN, International) expanded the partnership in January 2026. Commercial launch targeted 2027.
• Differentiator: Asset light, factory embedded model; avoids the capital burden of owning a fleet.
• Critical take: Going public via SPAC after the last AV SPAC wave busted (Embark collapsed) is a yellow flag on capital quality. Pre revenue with a 2027 launch, selling a promise into public markets, and dependent on partners like TRATON for timing and economics.

3. Drones and autonomous aviation

Skydio

• What: AI powered autonomous drones for defense, public safety, and infrastructure inspection.
• HQ: San Mateo, CA.
• Founded: 2014 by Adam Bry, Abraham Bachrach, Matt Donahoe (MIT and former Google Project Wing).
• Funding: Series F of $110M in early 2026 at ~$4.4B; ~$740M total; hundreds of millions in annual revenue.
• Traction: 3,800+ customers, 60,000+ drones delivered, ~600 public safety agencies. Major defense wins including a $52M US Army order (2,500 units).
• Differentiator: Best in class onboard visual autonomy plus a US made, NDAA compliant supply chain as China's DJI faces US restrictions.
• Critical take: Growth is increasingly tethered to defense budgets and to DJI restrictions holding. Funding a $3.5B US manufacturing buildout on "hundreds of millions" of revenue implies heavy dilution ahead.

Zipline

• What: Autonomous fixed wing and hovering "droid" delivery drones for medical supplies, food, and retail.
• HQ: South San Francisco, CA.
• Founded: 2014 (roots in Romotive, 2011) by Keller Rinaudo Cliffton, Keenan Wyrobek, and others.
• Funding: Series H around $800M ($600M in January 2026 at a $7.6B valuation, plus $200M in March 2026). Backers: Valor, Tiger Global, Fidelity, Baillie Gifford.
• Traction: 2M+ commercial deliveries; operating across five African countries, multiple US cities, and Japan; partners include Walmart and health systems.
• Differentiator: By far the most mature, highest volume autonomous delivery operator globally, with deep regulatory experience.
• Critical take: Clear leader, but US suburban expansion (Platform 2) must prove economics the way African medical corridors did. Restaurant and retail drone delivery is a noisier, lower margin use case than lifesaving logistics, and the valuation is near 4x the 2024 mark.

Reliable Robotics

• What: A retrofit autonomy system that flies existing aircraft (starting with the Cessna 208 cargo plane) with no pilot on board.
• HQ: Mountain View, CA.
• Founded: 2017 by Robert Rose and Juerg Frefel (both SpaceX alumni).
• Funding: $160M round in April 2026 near a $1B valuation; ~$300M total.
• Traction: 200+ orders; $17.4M US Air Force contract; selected for the FAA Advanced Air Mobility program.
• Differentiator: Pursuing full FAA type certification under existing rules (no special waivers), aimed at real airspace integration.
• Critical take: A long, certification gated bet. The type cert timeline already slipped from 2028 to 2029, and first of kind FAA timelines tend to slip further. Meaningful revenue is years away at a near $1B valuation.

Corvus Robotics

• What: Fully autonomous inventory drones ("Corvus One") that fly warehouse aisles for cycle counting, now hardened for freezers.
• HQ: Mountain View, CA (MIT roots).
• Founded: 2017 by Jackie Wu and Mohammed Kabir.
• Funding: ~$28M total ($18M Series A, September 2024). Smallest war chest in the delivery/drone group.
• Traction: Kroger using the cold chain freezer system in live operations.
• Differentiator: Fully autonomous indoor flight with no external beacons, navigating aisles as narrow as 50 inches down to -20F.
• Critical take: Competes against ground robots, fixed cameras, and cheaper manual counting. The cold chain niche is a smart wedge, but a single flagship customer (Kroger) is concentration risk for a thinly funded startup.

4. Sidewalk and last mile delivery

Starship Technologies

• What: Six wheeled autonomous sidewalk robots for last mile food and grocery delivery, concentrated on campuses.
• HQ: San Francisco, CA (major operations in Estonia).
• Founded: 2014 by Ahti Heinla and Janus Friis (Skype co founders).
• Funding: $50M Series C in October 2025; $280M+ total.
• Traction: 2,700+ robots, 9M+ deliveries, ~65 US campuses; new Uber Eats partnership; goal of 12,000+ robots by 2027.
• Differentiator: The largest and most operationally proven sidewalk fleet in the world.
• Critical take: A decade in, sidewalk delivery has never proven strong economics beyond enclosed campuses. The jump to open cities brings regulation, weather, and theft. A modest raise for a plan to quadruple the fleet.

Serve Robotics

• What: Autonomous sidewalk delivery robots, plus (via 2026 acquisition) indoor service robots.
• HQ: Redwood City, CA. Public (SERV).
• Founded: Spun out of Uber/Postmates; incorporated as Serve in 2021. Founders Ali Kashani and Dmitry Demeshchuk.
• Funding: $247M+ raised in trailing 12 months as of early 2025; ~$260M cash end of 2025. Nvidia is a shareholder.
• Traction: 2,000+ robots across 20 cities; ~3,600 restaurants; deep Uber Eats integration. FY2025 revenue $2.7M, 2026 guidance ~$26M.
• Differentiator: Largest US sidewalk fleet, Nvidia based autonomy, vertical integration.
• Critical take: Guiding to a 10x revenue jump while pausing new sidewalk deployments in H1 2026 is a red flag. The Diligent Robotics acquisition signals sidewalk delivery alone may not scale to the story investors are pricing. Effectively reliant on Uber Eats for demand.

Cartken

• What: Autonomous wheeled robots pivoting from sidewalk food delivery to indoor and outdoor industrial material handling.
• HQ: Oakland, CA.
• Founded: 2019 by former Google engineers (Christian Bersch and team).
• Funding: ~$20M to $22.5M total (Mitsubishi Electric strategic investor). Smallest and least capitalized here.
• Traction: A Mitsubishi Electric group company ordered nearly 100 Cartken Hauler robots (carries ~660 lbs).
• Differentiator: Camera first, all weather autonomy that bridges factory interior and outdoor yard movement.
• Critical take: A mid life pivot (announced mid 2025) means the original sidewalk thesis did not pencil out. Now entering a crowded industrial AMR field against better funded incumbents, with a possible single customer dependency.

5. Humanoids (hardware first)

Figure (Figure AI)

• What: General purpose humanoid robots plus an in house vision language action model (Helix).
• HQ: Sunnyvale, CA.
• Founded: 2022 by Brett Adcock (previously Vettery, Archer Aviation).
• Funding: ~$1.9B total; Series C above $1B in September 2025 at a $39B valuation (roughly 15x the early 2024 mark). Backers include Nvidia, Intel, Qualcomm, Salesforce.
• Traction: BMW Spartanburg pilot ran ~1,250 hours across 90,000+ parts. Figure 03 is the home and mass production platform; goal of 100,000 robots in four years.
• Differentiator: Vertically integrated, proprietary Helix model (moved off OpenAI to in house AI in 2024), own actuators and hands.
• Critical take: Valuation is wildly ahead of revenue; deployments are pilots, not paid at scale. The 100,000 unit target and home pivot are unproven, and marketing has run ahead of demonstrated autonomy.

1X Technologies

• What: NEO, a soft bodied bipedal humanoid designed to work inside homes.
• HQ: Palo Alto, CA (relocated from Norway in 2025); manufacturing in Hayward and Moss, Norway.
• Founded: 2014 as Halodi Robotics by Bernt Bornich; rebranded 1X in 2022. OpenAI is an early backer.
• Funding: ~$125M at an $820M valuation (early 2024); reportedly raising ~$1B near a ~$10B valuation late 2025 (in progress, unconfirmed).
• Traction: Pre orders at $20,000 outright or $499/month, deliveries targeted late 2026; plans several thousand test units in homes.
• Differentiator: Home safe design, lightweight compliant body and tendon driven actuators for close human contact.
• Critical take: The shakiest autonomy claims in the field. Early NEO demos leaned heavily on teleoperation, so "several thousand homes" may mean a human remotely watching inside your house. Real privacy problem, and selling an unproven robot at $20K before autonomy is solved is a steep bet.

Tesla Optimus

• What: General purpose humanoid, for Tesla factories first, trained partly off Tesla's vision AI stack.
• HQ: Corporate in Texas; Optimus production at Fremont, CA. Not an independent company.
• Founded: Program announced 2021; part of Tesla Inc. (Elon Musk).
• Funding: Internal; no standalone valuation.
• Traction: No external commercial deployments yet. Targets are aspirational (reported 50,000 to 100,000 units in 2026; ambitions toward ~1M/year capacity).
• Differentiator: Leverages Tesla's existing AI, manufacturing scale, and in house components; Gen 3 targeted below $20K at scale.
• Critical take: Chronic timeline slippage is the story. The V3 reveal and mass production have been pushed multiple times, and Musk himself called the ramp impossible to predict. Hand dexterity and teleoperation questions mean the 2026 targets deserve heavy skepticism.

The Bot Company

• What: A low, non humanoid mobile base with an arm and grippers for home chores.
• HQ: San Francisco, CA.
• Founded: 2024 by Kyle Vogt (former Cruise CEO), Paril Jain, Luke Holoubek.
• Funding: $150M seed then $150M Series B (March 2025) at $2B; reportedly raising ~$250M above a $4B valuation (October 2025, in progress).
• Traction: Early stage, no disclosed deployments. Traction is founder pedigree plus capital.
• Differentiator: Deliberately non humanoid, lower cost form factor for the home chore problem.
• Critical take: Enormous valuation with no shipping product, led by a founder whose last venture (Cruise) collapsed under a safety scandal. Home manipulation is the hardest unsolved problem in robotics, and secrecy makes the tech impossible to verify.

Sunday Robotics

• What: "Memo," an autonomous home robot trained to do chores like laundry and clearing tables.
• HQ: Mountain View, CA.
• Founded: 2024 by Tony Zhao (creator of ALOHA) and Cheng Chi (inventor of Diffusion Policy).
• Funding: $200M total; $165M Series B in March 2026 (Coatue led) at a $1.15B valuation, roughly a year after founding. Pre revenue.
• Traction: Data from 500+ real homes via a patented "Skill Capture Glove" (~10M episodes); 50 household beta targeted for late 2026.
• Differentiator: Founders behind two foundational robot learning works; a data moat from in the wild home manipulation data.
• Critical take: Textbook frothy bet: unicorn valuation, zero revenue, one robot not yet in a single home ~18 months after founding. General purpose home manipulation is the hardest unsolved problem in the field, and every predecessor has slipped timelines by years.

Mind Robotics

• What: Full stack physical AI (foundation models plus purpose built robots) for dexterous manipulation in manufacturing.
• HQ: Palo Alto, CA.
• Founded: 2025, spun out of Rivian as "Project Synapse" by RJ Scaringe (Rivian CEO).
• Funding: Over $1B total in under a year ($115M seed, $500M Series A in March 2026, another $400M led by Kleiner Perkins) at a reported $3.4B valuation.
• Traction: Rivian is anchor partner, shareholder, and live training environment. Essentially pre commercial outside Rivian.
• Differentiator: Uses proprietary high volume factory data from Rivian, a data moat most startups lack.
• Critical take: Extraordinary valuation for a one year old company with essentially one customer (Rivian, itself cash constrained). The EV assembly data may not generalize; this looks like founder halo and AI hype pricing ahead of validation.

6. Robot foundation models (the "brains")

Physical Intelligence (pi)

• What: Robot foundation models (pi-zero, pi-0.5) meant to control any robot hardware. Not a robot maker.
• HQ: San Francisco, CA.
• Founded: 2024 by Karol Hausman, Sergey Levine, Chelsea Finn, Brian Ichter, Adnan Esmail, Lachy Groom.
• Funding: $400M Series A (November 2024, Bezos led, at $2.4B), then $600M at $5.6B (November 2025, CapitalG led); ~$1B near $11B reportedly in talks early 2026 (unconfirmed).
• Traction: pi-zero open sourced February 2025; pi-0.5 showed open world generalization in unseen homes. Adoption via the research community and partners.
• Differentiator: Elite research team and a credible "GPT for robots" thesis with real open source releases.
• Critical take: Hardware agnostic means no control of the full stack and dependence on partners for real world data, while hardware players build competing in house models. Robotics lacks an internet scale dataset, so whether one model generalizes across all robots is still open. Minimal revenue at an $11B mark.

Skild AI

• What: "Skild Brain," a single model meant to control bodies it was never trained on.
• HQ: Pittsburgh, PA (note: NOT Bay Area).
• Founded: 2023 by Deepak Pathak and Abhinav Gupta (Carnegie Mellon faculty).
• Funding: $300M Series A (2024, ~$1.5B), then a $1.4B raise around January 2026 above a $14B valuation (SoftBank, Nvidia, Bezos).
• Traction: Reportedly $0 to ~$30M revenue within months of 2025 (company stated); pilots across security, warehouses, manufacturing, construction.
• Differentiator: Cross embodiment generalization and robustness, leveraging CMU self supervised learning heritage.
• Critical take: Valuation is roughly 400x to 500x on early, pilot heavy revenue. "Control any robot it has never seen" is far easier to demo than to make reliable at commercial scale. A software layer competing against both peers and vertically integrated hardware firms.

Genesis AI

• What: A universal robotics foundation model plus a synthetic data and simulation platform.
• HQ: Bay Area (Palo Alto) with a Paris presence (exact HQ unverified).
• Founded: December 2024 by Zhou Xian (CMU) and Theophile Gervet (former Mistral).
• Funding: $105M seed (July 2025, Eclipse and Khosla co led, Eric Schmidt among backers). No later round disclosed.
• Traction: Emerged from stealth July 2025; unveiled first model (GENE-26.5) around May 2026. Pre commercial.
• Differentiator: Heavy reliance on synthetic data from a proprietary physics engine to sidestep the real world data bottleneck.
• Critical take: The core bet, that simulation can substitute for real world data, is exactly where sim to real transfer has historically broken down. Least funded lab here against billion dollar rivals, plus possible China hardware supply chain friction.

Dyna Robotics

• What: DYNA-1, a single weight general purpose model aimed at commercially useful repeatable manipulation.
• HQ: San Mateo area, CA.
• Founded: 2024 by Lindon Gao and York Yang (sold Caper AI for ~$350M) plus Jason Ma (former DeepMind).
• Funding: $143.5M total ($120M Series A, September 2025; Nvidia, Amazon, Samsung, LG among backers).
• Traction: Claims 99+% success in 24 hour nonstop operation and early paying deployments (company stated).
• Differentiator: Focus on reliability and deployability now, backed by founders with a proven exit.
• Critical take: Smallest and least hyped of the labs, at risk of being out capitalized 10x by Skild and Physical Intelligence. The 99+% figures are on narrow repetitive tasks; the leap to the "physical AGI" it markets is enormous and unproven.

7. Manipulation, arms, and picking

Dexterity

• What: AI robotic arms for truck loading, palletizing, and singulation, now extended to an industrial "superhumanoid" (Mech).
• HQ: Redwood City, CA.
• Founded: 2017 by Samir Menon (Stanford robotics PhD).
• Funding: ~$291M total; $95M in March 2025 at a $1.65B valuation (Lightspeed, Sumitomo).
• Traction: FedEx and UPS deployments; long standing Kawasaki partnership; June 2026 expansion to scale Mech production.
• Differentiator: A "physical AI" stack running hundreds of coordinated models on onboard compute, dual arm coordination, force sensing.
• Critical take: Pivoting toward a capital intensive humanoid puts it in a crowded, hype heavy race (Figure, Agility) where duty cycle reliability is unproven. The valuation prices in humanoid success not yet demonstrated at scale.

Chef Robotics

• What: AI robotic arms that portion and assemble food in high mix production lines.
• HQ: San Francisco, CA.
• Founded: 2019 by Rajat Bhageria.
• Funding: ~$66M total ($43M Series A in April 2025, part equity part equipment debt).
• Traction: Amy's Kitchen, Sunbasket, Cafe Spice; 40M+ meals assembled; UK entry in 2025.
• Differentiator: Robotics as a service (no customer capital expenditure) plus AI handling variable, deformable food, with a data flywheel.
• Critical take: Food assembly is low margin and labor is cheap, so ROI must be tight. Heavy reliance on equipment financing debt for the service model is balance sheet risk. Still early (Series A) versus larger funded peers.

Nimble Robotics (warehouse fulfillment)

• What: A general purpose picking robot plus a fully autonomous 3PL fulfillment network.
• HQ: San Francisco, CA.
• Founded: ~2017 by Simon Kalouche.
• Funding: ~$236M+ total; $106M Series C in October 2024 at a $1B valuation, led by FedEx. No confirmed 2025 to 2026 raise.
• Traction: Strategic FedEx agreement; board includes Fei-Fei Li, Marc Raibert, Sebastian Thrun.
• Differentiator: Claims one general purpose robot doing all core fulfillment functions, delivered as an end to end autonomous 3PL.
• Critical take: Running its own fulfillment network is enormously capital intensive against Amazon and incumbent 3PLs. FedEx as both lead investor and anchor customer is concentration risk, and "fully autonomous" fulfillment is hard to verify externally.

Flexiv

• What: Adaptive force controlled robot arms (Rizon) blending force control with AI and vision.
• HQ: Santa Clara, CA (major operations in China).
• Founded: 2016 by a Stanford team (CEO commonly cited as Shiquan Wang; name spellings vary).
• Funding: Reported $122M to $322M across sources (figures conflict); unicorn ($1B).
• Traction: Automotive, electronics, and internet industry customers across China, North America, Europe, South Korea.
• Differentiator: Hierarchical force control plus AI and vision giving arms human like adaptivity to contact rich, imprecise tasks.
• Critical take: US headquarters but China centric operations create geopolitical and procurement risk (US customers wary). Conflicting funding data suggests limited transparency, and it competes on price against FANUC, UR, ABB.

Ambi Robotics

• What: AI robotic parcel sorting and induction (AmbiSort) with a soft touch gripper.
• HQ: Berkeley, CA.
• Founded: 2018 (as Ambidextrous Laboratories) by Ken Goldberg, Jeff Mahler and others, spun out of Berkeley's AUTOLab (Dex-Net).
• Funding: ~$65M total ($32M latest; Tiger Global, Pitney Bowes strategic).
• Traction: Pitney Bowes and OSM Worldwide (multi year service deal); 1,200+ items/hour at 99+% accuracy.
• Differentiator: Sim to real transfer learning plus a soft touch gripper for mixed rigid and deformable parcels.
• Critical take: Narrow use case (sortation) where labor is cheap and rivals (Covariant, now largely absorbed by Amazon) chase the same niche. Dependence on a few logistics customers makes revenue durability the key question.

Robust AI

• What: "Carter," a collaborative human guided mobile cart for warehouse picking and transport.
• HQ: San Carlos, CA.
• Founded: 2019 by Rodney Brooks (iRobot, Rethink Robotics co founder).
• Funding: ~$42.5M total. By far the leanest in this group. No large 2025 to 2026 mega round.
• Traction: DHL reference (60+% productivity gain); November 2025 partnership with Aptiv.
• Differentiator: A force sensitive, intuitive human robot interface; a software defined cart doing picking assist, transport, and mobile sort.
• Critical take: Materially under capitalized in a capital intensive category. Leaning on Aptiv suggests it cannot fund full stack scale up alone, and monetization beyond pilots is unproven.

Cobot (Collaborative Robotics)

• What: "Proxie," an autonomous mobile manipulator that moves carts in warehouses, hospitals, and manufacturing.
• HQ: Santa Clara, CA.
• Founded: 2022 by Brad Porter (former Amazon VP who scaled Amazon Robotics to 500k+ robots).
• Funding: Over $140M total ($100M Series B, April 2024). No confirmed newer round.
• Traction: Mayo Clinic, Maersk, Moderna; ~154,000+ carts moved across ~13,000 hours; a peer reviewed Mayo Clinic pilot (October 2025).
• Differentiator: Pragmatic wheeled mobile manipulator (avoids expensive humanoids) with strong situational awareness and founder operational credibility.
• Critical take: Proxie is closer to an AMR that pushes carts than a dexterous picker; the manipulation claims are still maturing. Crowded transport market (Locus, Fetch/Zebra) with thin differentiation, and funding data may be stale.

8. Warehouse, retail, and consumer automation

Mytra

• What: A modular 3D lattice storage system where 3,000 lb capable bots climb inside a cube structure.
• HQ: South San Francisco, CA.
• Founded: 2022 by Chris Walti (former head of Tesla's Optimus program) and Ahmad Baitalmal.
• Funding: $200M+ total; $120M Series C in January 2026 (Avenir led); reported ~$2B valuation (soft, unverified).
• Traction: Signed a Fortune 100 food company and a Fortune 500 distributor; added former Tesla CFO Zach Kirkhorn to the board.
• Differentiator: Patented cube design decoupling storage density from throughput; pitched as a full building operating system.
• Critical take: Crowded, capital intensive category dominated by AutoStore, Exotec, Symbotic. The $2B valuation looks priced for a category winner while customer count is a handful of unnamed enterprises.

Simbe Robotics

• What: "Tally," an autonomous shelf scanning retail inventory robot, plus fixed sensors.
• HQ: South San Francisco, CA.
• Founded: 2014 by Brad Bogolea, Jeff Gee, Mirza Shah.
• Funding: ~$100M+ total ($50M Series C, October 2024, Goldman led).
• Traction: ~29 retail brands (SpartanNash, BJ's, ShopRite); Tally 4.0 launched January 2026.
• Differentiator: The longest tenured shelf audit platform, a robot plus fixed sensor hybrid with an analytics subscription.
• Critical take: Retailers have killed in aisle robots before (Walmart dropped Bossa Nova). The pivot toward fixed sensors quietly concedes the roaming robot alone may not be the answer. Twelve years in with only ~$100M raised signals slow adoption.

Bear Robotics

• What: "Servi" hospitality service robots for restaurants and hotels, expanding into cleaning.
• HQ: Redwood City, CA.
• Founded: 2017 by John Ha (former Google engineer), Juan Higueros, Fangwei Li, Brennand Pierce.
• Funding: $173M+ raised. Now majority owned by LG (LG reached ~51% in January 2025).
• Traction: 16,000+ robots shipped, one of the highest deployed counts in service robotics.
• Differentiator: Large installed base plus LG's manufacturing scale, distribution, and capital.
• Critical take: LG control de risks capital but ends independence and subordinates Bear to LG's roadmap. The category is brutally commoditized by cheap Chinese OEMs (Pudu, Keenon) in a race to the bottom.

Matic Robots

• What: A vision only (no lidar) home robot that vacuums and mops in one pass.
• HQ: Mountain View, CA.
• Founded: 2017 by Navneet Dalal, Mehul Nariyawala (previously sold Flutter to Google), Shafaq Presswala.
• Funding: ~$107M total; ~$77M in July 2025 at ~$650M. Angels include Nat Friedman, the Collison brothers, Jack Dorsey.
• Traction: Direct to consumer shipping; manufacturing expansion in Mountain View. Unit sales not publicly verified.
• Differentiator: Vision first, on device autonomy for privacy; positioned as the premium "what comes after Roomba."
• Critical take: A $650M valuation is aggressive where iRobot has struggled and Chinese players (Roborock, Ecovacs, Dreame) dominate on price. The bet that vision plus AI justifies a flagship phone price when $500 alternatives are good enough is unproven.

9. Construction, surveying, and agriculture

Built Robotics

• What: Autonomous retrofit kits for heavy equipment, now focused on a solar pile driving system (RPD 35).
• HQ: San Francisco, CA.
• Founded: 2016 by Noah Ready Campbell (former Google).
• Funding: ~$112M total ($64M Series C, 2022). No new priced round since 2022.
• Traction: 40+ deployments; a ~$75M contract with Blattner Energy on a Louisiana solar farm powering a Meta data center.
• Differentiator: Pivoted from generic autonomy to a purpose built, vertically integrated solar piling robot.
• Critical take: Has raised far less than newer rival Bedrock and no round since 2022. Now a niche solar player tightly coupled to the solar build cycle and one big customer (Blattner concentration risk).

Bedrock Robotics

• What: A reversible same day retrofit that makes excavators and dozers fully autonomous.
• HQ: San Francisco, CA.
• Founded: 2024 by Boris Sofman (former Waymo trucking lead, Anki co founder) and an ex Waymo team.
• Funding: $350M+ total; $270M Series B in February 2026 (CapitalG and Valor); reported ~$1.75B valuation (secondary sources, unconfirmed).
• Traction: Completed a supervised autonomy mass excavation on a 130 acre site (~November 2025). Targeting operator out deployments in 2026.
• Differentiator: An elite ex Waymo team applying AV grade perception to construction, with deep Nvidia and Alphabet backing.
• Critical take: Massively funded but pre revenue scale and still at supervised autonomy. The whole thesis rests on hitting true operator out deployment in 2026, a milestone AV firms have repeatedly missed. Overlaps with Caterpillar and Komatsu, who already ship autonomy.

Dusty Robotics

• What: "FieldPrinter," an autonomous robot that prints digital building layouts (BIM) onto construction floors.
• HQ: Mountain View, CA.
• Founded: 2018 by Tessa Lau (former Savioke CTO) and Philipp Herget.
• Funding: $70M total; last priced round $45M Series B in 2022 ($250M valuation, now stale).
• Traction: Dozens of contractors (including Swinerton); 25M+ sq ft printed in a single quarter.
• Differentiator: First mover in robotic BIM to floor layout with millimeter accuracy; deepened by a Leica laser tracker integration.
• Critical take: Strong product fit, but ~3 years without a new raise while rivals raise hundreds of millions. Layout printing may be a feature that Hexagon or Trimble could absorb rather than a standalone market. Single product concentration.

Civ Robotics

• What: "CivDot," a small autonomous robot for high precision GPS land surveying and layout marking.
• HQ: San Francisco, CA (R&D in Tel Aviv).
• Founded: 2018 by Tom Yeshurun.
• Funding: $12.5M total ($7.5M Series A, July 2025, Bobcat strategic).
• Traction: 100+ robots deployed; reached profitability in 2025; marks up to 3,000 points/day.
• Differentiator: A small, rugged, low cost, already profitable point marking robot with Bobcat distribution backing.
• Critical take: Impressive capital efficiency, but very small and thinly funded against rivals raising 20x to 40x more. Risk of being bundled into larger autonomy platforms (Trimble, or excavator retrofit players), and dependent on the solar build out.

Agtonomy

• What: A physical AI software platform (TeleFarmer) embedded into tractors, one operator supervising 10+ machines.
• HQ: South San Francisco, CA.
• Founded: ~2020 to 2021 by Tim Bucher (former Apple, Microsoft; a working vineyard owner).
• Funding: ~$64M total ($18M Series B, October 2025; Bobcat and Kubota strategic).
• Traction: Kubota showcased an autonomous tractor running Agtonomy at CES 2026 and made a direct investment (April 2026); launched in Australia.
• Differentiator: Software first, OEM partnership model; the "brains" layer for existing farm and turf equipment.
• Critical take: The OEM embedded strategy is double edged: Kubota and Bobcat could build or buy autonomy in house and relegate Agtonomy to a commodity supplier. Modest funding for capital intensive field autonomy, a graveyard of well funded failures.

10. Security

Knightscope

• What: Autonomous security robots (the K5) plus emergency call boxes, and now human guarding services.
• HQ: Mountain View, CA. Public (KSCP).
• Founded: 2013 by William Santana Li (former Ford) and Stacy Dean Stephens.
• Financials: FY2025 revenue $11.3M (+5%), net loss $33.8M, cash $20.6M. Q1 2026 revenue $6.0M (+106% YoY) driven by an acquisition. Micro cap with a history of reverse splits and dilution.
• Traction: Robots and call boxes across US hospitals, campuses, malls, and government sites. February 2026 acquisition of Event Risk added a 400+ person human guarding workforce.
• Differentiator: The only pure play publicly traded security robot company, now a hybrid of robots plus human guards.
• Critical take: After 12+ years the core robot business is deeply unprofitable, and 2026 "growth" comes from buying a low margin human guarding company, which contradicts the original robots replace guards thesis. Watch cash runway and going concern risk.

11. Platforms, software, and sensors (picks and shovels)

NVIDIA robotics

• What: The full physical AI stack: simulation (Omniverse, Isaac Sim), on robot compute (Jetson Thor), and open humanoid foundation models (Isaac GR00T).
• HQ: Santa Clara, CA. Public (NVDA), market cap ~$4.5T.
• Financials: Automotive and robotics segment ~$2.3B for the year (a rounding error against ~$44B/quarter of data center revenue).
• Traction: Isaac and GR00T adopted across the humanoid ecosystem (Figure, Agility, Boston Dynamics, 1X); CES 2026 releases (Cosmos, GR00T N1.6, Jetson T4000).
• Differentiator: Owns the entire loop: synthetic data and sim to real training, the foundation model, and the inference silicon. No competitor spans sim plus model plus chip.
• Critical take: GR00T is a loss leader to sell chips, and its models remain research grade. No one has shown a GR00T powered robot doing reliable, economically useful work at scale. NVIDIA is selling shovels for a gold rush that has not proven there is gold.

Ouster

• What: High resolution digital lidar sensors for automotive, robotics, and infrastructure.
• HQ: San Francisco, CA. Public (OUST). Merged with Velodyne in 2023.
• Financials: Q1 2026 revenue $49M (+49% YoY); raised $200M in July 2026 to scale Rev8 color lidar.
• Traction: Smart infrastructure is the story (700+ traffic deployments, 550+ security sites); qualified for Nvidia DRIVE Hyperion.
• Differentiator: Digital lidar (single custom chip vs discrete analog parts), lower cost and Moore's law style scaling; Rev8 marketed as first native color lidar.
• Critical take: Still burning cash and just diluted at a high price. The highest margin slice (Velodyne royalties) evaporates in 2026, so margins likely compress in a price deflationary market (Hesai, Aeva). The bull case depends on robotics and AV volumes that keep getting pushed out.

RealSense

• What: Stereo depth (3D) cameras and vision SDK for robots, AMRs, and humanoids.
• HQ: Santa Clara, CA.
• Founded: Intel's RealSense group, spun out as independent RealSense Inc in 2025 (CEO Nadav Orbach).
• Funding: $50M Series A (announced July 2025; Intel Capital and MediaTek participating); dormakaba took a minority stake in November 2025.
• Traction: Claims its cameras are in ~60% of the world's AMRs and humanoids (company figure, unverified).
• Differentiator: A mature, widely adopted depth platform with a long standing SDK and ecosystem lock in.
• Critical take: An incumbent cut loose right as depth sensing fragments (Luxonis, Orbbec, solid state lidar). The "60%" claim reflects a legacy Intel install base that may not survive without Intel's scale. $50M is thin for global manufacturing, and the pivot toward biometric access control hints pure robotics may not be lucrative enough alone.

Formant

• What: Cloud robot fleet management and operations software for mixed robot fleets.
• HQ: Mill Valley, CA.
• Founded: 2018 by Jeff Linnell (former Google robotics lead).
• Funding: ~$45M total; last raise a ~$21M Series B in October 2023 (BMW i Ventures led). No 2025 to 2026 round found (possible funding gap).
• Traction: SoftBank Robotics America named customer; claims it powers tens of thousands of production robots (self reported).
• Differentiator: A hardware agnostic single pane of glass for mixed fleets, plus a newer AI ops layer.
• Critical take: Selling shovels into a market that has not scaled; enterprise robot fleets are still small. Squeezed by OEMs building their own software and Nvidia bundling free tooling into the same layer. Vendor agnostic is a virtue only if customers do not standardize on one vendor.

Foxglove

• What: A multimodal data and observability platform for robotics and physical AI teams.
• HQ: San Francisco, CA.
• Founded: 2021 by Adrian Macneil and Roman Shtylman (former Cruise infrastructure engineers).
• Funding: ~$58M total; $40M Series B in November 2025 (Bessemer led).
• Traction: Tens of thousands of developers; customers include Nvidia, Amazon, Anduril, Wayve, Dexterity.
• Differentiator: Brings the in house data infrastructure only giants could previously build to the whole industry as an off the shelf, vendor agnostic platform.
• Critical take: Classic dev tool monetization gap; much usage is open source and free tier, and converting developers into paying enterprise revenue is unproven. Risk that large customers build in house, and fully levered to physical AI actually reaching scale.

Intrinsic

• What: Industrial robotics software (Flowstate), the "Android of robotics" thesis.
• HQ: Mountain View, CA (also Munich).
• Founded: 2021 as an Alphabet "Other Bet," spun out of X; founding CEO Wendy Tan White. Alphabet owned.
• Funding: None external; wholly owned by Alphabet.
• Traction: October 2025 joint venture with Foxconn for electronics manufacturing automation. In February 2026, folded from an independent Alphabet bet into Google itself, working with DeepMind.
• Differentiator: A hardware agnostic software layer abstracting away low level robot programming, backstopped by DeepMind (Gemini) and Alphabet's balance sheet.
• Critical take: After ~5 years it still lacks a large commercial footprint, and the absorption into Google reads as much a consolidation signal as an upgrade. Alphabet has a long history of quietly killing bets that do not scale.

OpenMind

• What: OM1, an open source hardware agnostic "Android for robots" OS, plus FABRIC, a decentralized blockchain based coordination and machine identity protocol. This is the web3 crossover on the map.
• HQ: San Francisco / Palo Alto, CA.
• Founded: 2024 by Jan Liphardt (Stanford professor).
• Funding: $20M round in August 2025 led by Pantera Capital (Coinbase Ventures, DCG). An associated ROBO token exists; a cited ~$400M valuation is speculative crypto press framing (unverified). Round label (seed vs Series A) is inconsistent across sources.

Source: "Robotics Maps by Lukas M. Ziegler"