Last year, the company that arguably made the biggest splash at the North American International Auto Show wasn’t even a company that made cars. Waymo, the self-driving unit of Google parent Alphabet, came to Detroit with its autonomous minivan in tow with the unspoken intent of striking fear in the hearts of the auto industry. According to a new report released today, it’s working.

Waymo didn’t bother to show up to the Detroit Auto Show press days this year, but the shadow it casts is long. The Silicon Valley company has surged passed legacy car companies like Ford, Daimler, and Renault Nissan to grab the No. 2 spot in Navigant Research’s annual autonomous driving scorecard. General Motors is still in the lead, but Larry Page and Sergey Brin’s self-driving spinoff is nipping at the auto giant’s heels. (This is despite its complete inability to manufacture a car at scale.) Hey Google, not too shabby.

Since 2015, Navigant has scored the 20-or-so companies working on self-driving technology on 10 different criteria related to strategy, manufacturing, and execution. Basically, it asks, how good is their technology? Can they manufacture it at scale? And what’s their plan for getting it to the masses? After that, Navigant ranks the companies in four categories: leaders, contenders, challengers, and followers.

Last year, only four companies (and partnerships) landed in the leader category: GM, Ford, Daimler, and Renault-Nissan. This year, there are twice as many leaders: GM, Waymo, Daimler-Bosch, Ford, Volkswagen, BMW-Intel-FCA, Aptiv (née Delphi), and Renault-Nissan. The appearance of Waymo, Intel, Bosch, and Aptiv is a sign of the growing prominence of tech firms and auto suppliers in the race to build self-driving cars. Meanwhile, a growing bloc of the auto industry not only sees autonomy as inevitable, but is making real progress on it for their own cars.

GM continues to lead the pack, thanks to its willingness to spend a lot of money and make fairly insane-sounding promises. GM gets points for acquiring AV-oriented startups like Cruise Automation and LIDAR firm Strobe, expanding its car-sharing service Maven, and laying down clear benchmarks like launch a fleet of robot taxis without steering wheels and pedals in 2019. Most importantly, GM has the capability to mass produce its autonomous Chevy Bolts at scale. Navigant says the auto giant is “well positioned to have a successful early deployment of highly automated driving in the coming years.”

But Waymo has also made a series of bold moves in 2017 that have spurred its rise in the rankings. Partnerships with companies like Fiat-Chrysler, Lyft, and Avis help address Waymo’s main weakness: its lack of experience in building, maintaining, and owning a fleet of vehicles. Also, its plan to launch a fully driverless mobility service in Arizona in early 2018, making it the first to remove the human safety driver, earned it a surplus of points in the scorecard.

Navigant goes one step further and predicts that the first autonomous car you’ll likely ride in will be from either GM or Waymo. It is certainly the case that there is a growing rivalry between the 109-year-old legacy automaker and Google spinoff that’s a little over a year old. It will be an interesting fight, especially as more and more highly automated vehicles start shuttling real people around.

Two other big names that don’t appear to be making as much progress are Apple and Tesla. First, the former. Navigant notes that while the Cupertino-based tech giant has “never developed a product as complex as an automobile,” Apple has “existing capabilities that make it uniquely positioned to participate in the automated driving space.” The company has applied for a license to test autonomous vehicles in California, and its test vehicle has been spotted a few times out in public. There’s no guarantee an autonomous Apple car will become a reality, but with a market cap of $900 billion, the company can’t be easily dismissed.

And then there’s Tesla. Navigant assessment of the electric car phenomenon is sure to anger Elon Musk’s vociferous fanbase, but the truth hurts sometimes. It’s especially interesting because Tesla has been the most outspoken among automakers in terms of promising to deliver Level 5, fully autonomous vehicles in the near future. (Most companies purposefully avoid making any promises about the ability to buy these self-driving cars.)

But Tesla is over-promising and underperforming, Navigant argues:

In a May 2017 TED talk, Musk claimed the systems being built today would be Level 5 capable by 2019. However, this is unlikely to ever be achievable since Level 5 is defined as the car being able to operate without human intervention in all conditions. Current Tesla hardware lacks the ability to keep sensors clean and unobscured in poor weather as well as most of the redundant systems needed for fully automated driving, not to mention the lidar that most people in the field believe is necessary for highly automated vehicles. Even Nvidia has expressed doubt that the computing hardware it sells to Tesla is capable of supporting full automation reliably.

Navigant also cites Tesla’s struggle to scale up production of its Model 3, as well as persistent quality and reliability issues with the new mass-market EV. It docks the automaker for a poor ability to execute a sustainable business model. (Translation: its finances are a mess.) “Great vision is unlikely to be adequate without fundamental changes in the way the company operates,” the research firm concludes.