Teardown

Daily digest · 2026-09-11

Scan #051: four autonomy bets whose cash runway is now the story — Nuro's post-Uber-Lucid pivot from a $2.13B raise to a licensing longshot, Dexterity's $1.65B March-2025 mark on ~$21M ARR, AiDash rolling into Schneider's One Digital Grid on a $350M EV, and Built Robotics' Blattner/Quanta solar-piling bet as Cat and Komatsu ship native OEM autonomy — against four US incumbents whose 2026 numbers just re-priced the franchise: Big 5 Sporting Goods taken private at $1.45/share on 2 October 2025, Heartland Express's operating ratio blown out to 107.1% in FY25 after the $525M CFI deal, Eversource losing $1.6B pre-tax on offshore wind while Connecticut's PURA and Massachusetts DPU cut its rate case, and the day's one well-positioned exception — Travelers Companies still printing an 83.6% Q2 2026 combined ratio and 24.9% core ROE through 15,000 independent agents

Eight companies where the compounding is either about to stop or about to reprice. Four emerging autonomy/AI companies (Nuro, Dexterity, AiDash, Built Robotics) each carry a specific mechanism-level question about whether their wedge outlasts their cash runway before an OEM, hyperscaler, or acquirer bundles it away. Four incumbents (Big 5, Heartland, Eversource, Travelers) — three at-risk and one well-positioned — show what happens when capex-heavy operational moats meet a 2026 rate environment, a hostile regulator, or a shift in retail footprint economics.

Today’s eight sit on either side of the same question: when does an operating model stop compounding? The four emerging companies are each a specific bet on autonomy, machine-learning inference, or retrofit-tech capturing a moat before an OEM or a hyperscaler bundles it away — and each of them has a cash-runway or acquirer-bundle question that the 2026 funding market is now forcing them to answer. The four incumbents are the mirror: a $1.45/share take-private of a 400-store retailer, a truckload carrier whose $525M acquisition doubled top-line but blew its operating ratio from 84% to 107% in three years, a regulated utility bleeding out $1.6B on offshore wind while its regulator withholds its rate case, and — as the day’s one well-positioned counterexample — a P&C franchise still printing 83.6% CRs through 15,000 independent agents that no direct-to-consumer entrant has been able to displace.

Nuro — Logistics · Emerging. 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 cut ~30% of staff in May 2023, killed its custom R2 delivery vehicle program in 2024, and re-emerged as a licensing partner for the Uber–Lucid Gravity robotaxi program. Nuro has never disclosed revenue, still burns SoftBank/T. Rowe-era capital, and now depends on the Nvidia DriveOS and Mobileye Chauffeur stacks not being “good enough” for the same OEMs it needs as customers. The delta between the company’s peak headcount (~1,500 in Nov 2022) and today’s inferred ~800 tells the same story as the closed-out Kroger/Domino’s/7-Eleven partnerships from 2020-22 — everything shipped so far has been unwound.

Dexterity — Supply chain · Emerging. Redwood City “physical AI” robotics company that raised $95M in March 2025 at a $1.65B post-money, on an estimated ~$21M 2025 ARR (getlatka) — a ~75x forward multiple that only makes sense if the Mech dual-armed loader and the Foresight world model produce measurable per-station cost advantages over Amazon-acquired Covariant (Aug 2024). The Sumitomo/Sagawa 1,500-robot Japan deployment promised in 2024 has not been publicly counted; the Boston Dynamics Stretch installed base is 1,000+ units. Every incumbent-shipped autonomy stack (Symbotic, Amazon, Ocado, MHS) is competing on TCO Dexterity can’t match — the question is whether its RaaS unit economics turn positive before the next raise.

AiDash — Energy · Emerging. Palo Alto satellite-plus-AI vegetation-and-asset-risk platform for electric utilities, founded 2019 to attack the post-Camp Fire wildfire-liability problem. Raised $91.5M across four rounds (Series C-2 $58.5M Nov 2024, Lightrock/Shell Ventures/SE Ventures/National Grid Partners); reportedly on the block or in strategic-partnership talks with Schneider Electric at a ~$350M EV as it moves under Schneider’s One Digital Grid initiative. That EV pins the AI-utility SaaS multiple at ~2.5-3.5x ARR — signal that AI’s ability to charge on top of a Copperleaf/SAP IS-U asset-management workflow is bounded. Overstory ($14M seed), Ecopia AI and Planet Labs sit on the flank; Schneider’s own bundling is now the wedge that decides whether AiDash’s cross-utility neutrality survives ownership.

Built Robotics — Construction · Emerging. San Francisco autonomous heavy-equipment startup that raised $112M across NEA/Founders Fund/Tiger Global 2016-2022, then cut ~25-30% of staff across two 2023 rounds. Pivoted from a general-purpose Exosystem retrofit kit to solar-farm piling — a September 2025 Blattner/Quanta deal is the only recent commercial signal. Caterpillar (Cat MineStar), Komatsu (FrontRunner), John Deere and Volvo Autonomous Solutions are all shipping native OEM autonomy on the same excavator and dozer frames Built retrofits, while Trimble + Mincon are moving into solar piling as an integrated OEM package. The company is now a specialised solar-piling contractor with an aftermarket kit as its residual moat.

Travelers Companies — Insurance · Incumbent, well positioned. $65B-market-cap, 172-year-old top-tier US P&C franchise built around independent-agency distribution and disciplined commercial underwriting, printing an 83.6% Q2 2026 combined ratio and 24.9% core ROE while returning $4.2B to shareholders in the trailing twelve months. The day’s one well-positioned incumbent because the moat is documented, not assumed: 15,000 independent agents that no D2C entrant has been able to displace at scale, a Business Insurance and Bond & Specialty book that MGAs and insurtechs have chipped at without unseating, and a Personal Lines segment that Travelers has deliberately shrunk in CA and FL rather than write bad risk. The 20-year dividend record and quantum-2.0 pricing engine confirm the compounding is still working — the fair critique is that it is compounding slowly in a Personal Lines market where Progressive continues to take direct share.

Big 5 Sporting Goods — Retail · Incumbent, at risk. El Segundo-based ~414-store western-US neighborhood sporting-goods chain taken private on 2 October 2025 by a Worldwide Golf / Capitol Hill Group partnership at $1.45 per share — a ~$112.7M enterprise value on a company that did $1.06B in FY19 revenue. An 18-year dividend eliminated in 2024, a Nasdaq minimum-bid notice in 2024, four consecutive years of comp-store declines from -3% to -15%, and the final $1.45 take-out is the market’s answer to the 400-store, closeout-buying, warehouse-format neighborhood-store model in an era when DICK’S House of Sport is opening 100k-sqft experiential formats and Amazon/Fanatics/SHEIN eat the low end. The going-private is the exit, not the fix; whichever piece of the footprint Worldwide Golf recycles into golf retail, most of the store base is dead weight.

Heartland Express — Logistics · Incumbent, at risk. North Liberty, Iowa dry-van truckload carrier whose $525M all-cash CFI acquisition from TFI International on 31 August 2022 pushed the fleet from ~4,300 to ~6,320 tractors and ~$1.3B run-rate — then rode the freight recession from an 85% operating ratio in FY21 to 107.1% in FY25. Q2 2026’s 91.0% OR is a recovery but still worse than Knight-Swift, Werner and JBHT peers over the same period; the Gerdin family’s 40+ year discipline was to shrink into a downturn, and the CFI deal broke that discipline at the peak of the cycle. Underlying: dry-van rate per mile stuck at ~$1.85, industry driver turnover 89%+, autonomous trucking (Kodiak, Waabi, Aurora) collapsing the pilot count into commercial lanes over 2026-2027. Heartland’s fleet age target (~2 years) is still an advantage — but it is a driver-retention advantage that autonomous OEMs are about to obsolete.

Eversource Energy — Energy · Incumbent, at risk. New England’s largest energy delivery company (~$28-30B rate base, 4.4M customers in CT, MA, NH). In three years: torched ~$1.6B pre-tax on the aborted offshore-wind program (South Fork, Revolution, Sunrise stakes sold to GIP at a loss in 2024), took six Moody’s downgrades and an S&P notch cut since 2022, absorbed $500M+ in disallowed storm costs from Connecticut’s PURA, and had its most recent CT rate case pared down to ~13% of ask. October 2025 saw the political fight with PURA chair Marissa Gillett go public. The 25-year dividend record is intact but growth guidance (5-7% EPS) is now capex-constrained by regulator hostility, while data-center load in MA/NH is being routed around Eversource by direct C&I VPPAs. The rate-of-return-on-rate-base model works if — and only if — the regulator lets you earn on the rate base you’re building.

Full pages linked above. Yesterday’s scan: Scan #050.

Full deep dives