Waymo leads robotaxi confidence index; ASI and SoftBank sign JV as consumer study lands

The Road to Autonomy Robotaxi Index, refreshed 12 hours ago, puts Waymo on 86.7, up 2.7 on the week, ahead of Baidu Apollo Go on 80.9 and WeRide on 75.3. The composite Robotaxi Confidence Index sits at 57.5, down 0.1.

That is the newest hard number in the dossier, and it frames everything else in this week's autonomous driving news. An outside outfit runs the scoreboard, and it is not an audited financial disclosure. Read the ranking as sentiment, not revenue. The spread still matters: across the 40-odd names on the list, scores run from 86.7 down to 2.9.

Consumer appetite grows, use does not

CBT News reported 12 hours ago on a Cox Automotive study that found 55% of Americans are open to riding in a robotaxi, yet only 18% have. Awareness is 70% nationally and 89% in cities where autonomous services operate. In those cities, usage rises to 33%, meaning one in three residents has taken a ride. For 61% of non-users, the top barrier is discomfort with having no human driver, the study says. The same research, the 2026 Cox Automotive Autonomous Vehicle Study, marks ten years since the company first measured attitudes toward the technology.

Other numbers from it: ownership of Level 2 and Level 3 vehicles nearly tripled in a decade, from 12% to 34%. Consumer familiarity with AVs rose from 40% in 2016 to 58% today.

Interest in a fully driverless Level 5 vehicle fell from 17% to 10%. Level 4 remains the most appealing option at 28%, down from 30% in 2016. And 46% expect fully autonomous vehicles in their lifetime, up from 37% in 2021.

According to Vanessa Ton, Senior Manager of Market and Customer Research at Cox Automotive, safety concerns and high-profile incidents have slowed progress, while growing acceptance of robotaxis suggests attitudes are changing.

Note the direction of travel. The same report that shows familiarity rising also shows the public cooling on the most extreme form of autonomy. Buyers want the option to drive, and that is not the pitch a robotaxi operator makes.

SoftBank puts $225 million behind construction autonomy

On Wednesday, The Robot Report carried news that Autonomous Solutions Inc., or ASI, and SoftBank Group Corp. announced a joint venture to commercialize brand-agnostic autonomous heavy equipment. SoftBank has provided $225 million in new funding to ASI as part of the collaboration. ASI was founded in 2000 and is based in Mendon, Utah. Its Mobius system orchestrates mixed fleets of haul trucks, dozers, loaders and compactors, and the company says it already runs in production across heavy construction, agriculture, logistics and landscaping for Fortune 500 customers.

Mel Torrie, CEO of ASI, told The Robot Report that large infrastructure projects are being built at a pace and complexity traditional labour-intensive methods struggle to match, and that skilled equipment operators are hard to find.

"This joint venture with SoftBank positions ASI at the forefront of autonomous construction innovation, built for scale, engineered for longevity, and open to every major equipment brand on the jobsite," Torrie said.

Torrie is scheduled to speak at the 2026 RoboBusiness on 20 October. The JV's stated focus is developing and commercializing autonomous construction equipment for large infrastructure projects, with the money going toward growing ASI's commercial operations. That is a different market from robotaxis, with different regulation and a customer base of contractors rather than passengers. On this week's evidence, it is also easier to sell.

AstroForge takes the same bet into orbit

SpaceNews reported on 30 September that space mining company AstroForge plans to fly a spacecraft next year to test whether it can operate fully autonomously using artificial intelligence. The company has not named a launch date beyond 2027.

Strip out the setting and the pattern is familiar: a company that cannot hire enough qualified operators, or cannot put a human in the loop at all, moves the decision-making into software. Space is the cleanest version of that argument, because there is no driver's seat to argue about.

The index's losers and the politics ahead

Back on the index, the seven-day moves are lumpy. Neolix gained 10.5 points to 66.4, Momenta added 8.0 to 53.9 and Kodiak 6.1 to 54.0. Waymo's 2.7 gain is small in percentage terms, but it is the highest absolute score on the board by 5.8 points. On the other side, Didi Autonomous Driving fell 8.1 to 47.7, DoorDash Dot lost 5.6 to 48.5, Aurora dropped 2.4 to 51.5 and Bot Auto slipped 2.5 to 36.5. Tesla sits at 50.5, up 1.7.

The index page also lists commercial tie-ups by platform, separating Uber, Lyft, Grab and Bolt relationships into exclusive, market-exclusive and non-exclusive categories. It does not say which companies hold which, so read the list as a taxonomy rather than a scorecard of deals.

Regulatory pressure is building in parallel. Recent headlines, noted here for context rather than as sourced fact, include a Minneapolis council proposal to mandate human drivers in robotaxis, reported by Minnesota Reformer on 30 September, and a Chinese draft law that would require automakers or importers to handle traffic violations committed in fully autonomous mode. Meanwhile the deployment news keeps coming. Waymo is reportedly preparing Berlin tests ahead of a 2027 launch and a Munich launch by the end of 2027, per EVMagz's autonomous category page. Zurich Airport has begun running two fully automated electric shuttles with WeRide, without onboard drivers or security personnel. Wayve and Uber have launched a supervised autonomous ride service in London. Kodiak AI is targeting unsupervised long-haul driverless service between Dallas and Houston by the end of 2026, and Aurora has said it expects more than 30,000 self-driving trucks generating $5 billion in revenue by 2030.

How much of that lands is the open question. The Cox data suggests the constraint is not curiosity but exposure: 55% say they would ride, and only 18% have had the chance. Every city that opens a service moves that second number, and the first one is already banked.