Urban Autonomy Lab
Chapter 2 · EV103

GM Super Cruise Crosses 1 Billion Hands-Free Miles; GM Confirms Cadillac ESCALADE IQ Eyes-Off Launch in 2028 — 28 April / 1 May 2026

Date collected: 11 May 2026
Summary

750,000 vehicles, 28.7M trips/year personally-owned ADAS — operationally dominant over robotaxi mode in US; Cruise-to-Super-Cruise pivot as Legitimacy phase 5

Headline Finding

On 28 April 2026 (corporate release) and reaffirmed in industry coverage 1 May 2026, General Motors announced that customers have collectively driven 1 billion hands-free miles using its Super Cruise driver-assistance system. The milestone covers nearly 750,000 Super Cruise-equipped vehicles across 23 GM models in North America. In the past year, GM customers used Super Cruise for 7.1 million hours across 28.7 million trips, with Super Cruise vehicle volume up ~70% and user count up ~80% year-over-year. GM concurrently confirmed that in 2028 it will launch eyes-off automated driving on the Cadillac ESCALADE IQ — a Level 3-style capability that allows the driver to disengage attention on limited-access highways. Last month, GM’s next-generation automated technology — trained on millions of real-world miles and validated in simulation — began supervised public-road testing on limited-access highways in California and Michigan, with more than 200 development vehicles entering live traffic. The 1B-mile milestone is GM’s strongest empirical signal yet that personally-owned Level 2+ ADAS is now the regulatory and market-volume centre of US automation — not robotaxis. For the CAM paper, the Super Cruise volume + Cadillac 2028 eyes-off pathway is the operational basis for asking whether the paper’s Ch.5 Failure Mode framework adequately treats personally-owned ADAS as distinct from ADS robotaxi deployment.

Detail

  • Announcement: 28 April 2026 GM corporate release; covered Repairer Driven News, Autoblog, Seoul Economic Daily 1 May 2026.
  • Milestone: 1 billion cumulative customer hands-free miles on GM Super Cruise. Equivalent to ~2,100 round-trips Earth-to-Moon.
  • Equipped vehicle base: Nearly 750,000 Super Cruise-equipped vehicles across 23 GM models (Cadillac, GMC, Chevrolet brands).
  • Past-year usage: 7.1 million Super Cruise hours; 28.7 million Super Cruise trips. Vehicle count up ~70%; user count up ~80% YoY.
  • Eyes-off plan:
    • 2028 launch: Cadillac ESCALADE IQ — first GM consumer model with eyes-off Level 3-style operation on limited-access highways.
    • Architecture: Trained on millions of real-world Super Cruise miles plus simulation. Designed to scale across ICE and EV platforms.
    • Current testing: Last month (April 2026), more than 200 development vehicles began supervised public-road testing on limited-access highways in California and Michigan.
  • Affordability framing: GM positioning the 2027 Chevrolet Bolt as “the most affordable way to go hands-free in America,” extending Super Cruise availability into the mass-market price segment.
  • Cruise legacy: GM’s robotaxi unit Cruise was wound down in December 2024 after the October 2023 San Francisco pedestrian-dragging incident. GM revived AV efforts in mid-2025, redirected from robotaxi services to personally-owned vehicles.

Governance Implication

Three points relevant to the paper:

  1. Personally-owned Level 2+ ADAS is the dominant operational reality of US automation in 2026, not robotaxis. Waymo currently runs ~500,000 fully driverless paid rides per week across 10 US cities. Tesla Robotaxi runs supervised and unsupervised services across three Texas cities with a fleet of 25–35 vehicles. Avride, Motional, Pony.ai operations are smaller still. GM’s Super Cruise, by contrast, is operational on 750,000 personally-owned vehicles, used 28.7 million times annually. The paper’s Ch.5 Failure Modes framework — built around fleet-coherent failures and city-level Allocation/Demand dynamics — should explicitly treat personally-owned ADAS as a distinct (not lesser) failure-mode surface. Allocation, Demand, Fiscal, Energy, and Legitimacy failure modes apply differently to a 750,000-vehicle, personally-owned, single-OEM ADAS fleet than to a multi-thousand-vehicle, operator-owned, multi-OEM robotaxi fleet. The Ch.5 chapter should make this distinction structurally visible.

  2. The 2028 eyes-off launch on Cadillac ESCALADE IQ is a Ch.9 regulatory horizon worth naming. Eyes-off Level 3 automation on limited-access highways requires either UNECE R157 Automated Lane Keeping System extensions or an equivalent federal-state regulatory framework in the US. The UNECE WP.29 / GRVA June 2026 vote (EV52) is on the broader ADS GTR — which would include eyes-off Level 3 capability. The Ch.9 Regulatory Horizon should treat 2028 as the year the institutional framework for eyes-off operation on personally-owned US vehicles is operationally needed, not just for robotaxi services. The institutional ODD framing of the paper applies as much to personally-owned eyes-off vehicles as to robotaxis — but with different governance interlocutors (state DMVs, NHTSA, OEM safety cases, insurance providers) and different failure modes (driver-attention-failure, ADAS-handoff failure, distinction-by-feature-set rather than geographic ODD).

  3. The Cruise → Super Cruise migration is a worked example of regulatory rerouting in response to a Legitimacy collapse. GM’s December 2024 wind-down of Cruise — after the San Francisco pedestrian-dragging incident — represented the first major US AV operator-collapse in the paper’s Legitimacy failure-mode framing. The Super Cruise volume milestone now demonstrates that GM has rerouted its automation investment from robotaxi-direct deployment into personally-owned ADAS. The paper’s Ch.5 Legitimacy chapter has named four phases of legitimacy life-cycle (anticipatory friction, operational legitimacy challenge, polity-level refusal, operator-collapse). The Cruise-to-Super-Cruise pivot should be treated as a fifth phase: operational redirection — an operator-collapse in one product mode followed by capability redeployment in a different product mode that bypasses the failed regulatory channel. The paper should treat this as a Legitimacy-mode resolution channel worth naming alongside the four prior phases.

Sources

Thread