Welcome back to TechCrunch Mobility, your hub for the future of transportation and now, more than ever, the role AI is playing in it. To get this in your inbox, sign up here for free — just click TechCrunch Mobility!

I’m back from a sojourn through New Hampshire’s White Mountains — a few days of respite before Disrupt 2026, TechCrunch’s annual tech conference in San Francisco. Now, after a few days immersed in fall colors, I am ready for the action. And there is gonna be a lot of it. 

I want to see you all there, so here is a 30% discount on tickets to Disrupt with code mobility30. For more info on the show, scroll down.

Now, on to the news!

Self-driving truck companies Aurora Innovation and Kodiak AI received a federal exemption this week that removes a massive barrier to commercialization. 

Under federal regulations, if a traditional big rig runs into trouble, the human driver must pull over, activate the hazard lights, and then physically place reflective warning triangles on the road within 10 minutes to alert other road users. But if a mechanical problem forces a self-driving truck to pull over, there’s no human driver to deploy those warning devices.

It was a big enough hurdle that Aurora took federal safety regulators to court over the requirement. After the court denied its request for an exemption, the company escalated the fight to the District of Columbia Court of Appeals. (Companies did have a temporary waiver as it awaited a decision.)

Now Aurora and other self-driving truck developers have been given the green light from the Federal Motor Carrier Safety Administration. The agency has granted a five-year exemption allowing companies to replace roadside warning devices, such as reflective warning triangles, with cab-mounted warning beacons. Check out Aurora’s first responders page to see how it works.

Daniel Goff, vice president of external affairs at Kodiak AI, said the exemption will help the industry “usher in an autonomous era of freight movement on U.S. roads, one that can save lives and improve the efficiency of goods delivery.”

Gerardo Interiano, Aurora’s head of government relations and public affairs, echoed that sentiment, saying the decision underscores the government’s recognition of the economic and community benefits of autonomous trucking. He described the high-visibility, cab-mounted warning beacons as a “critical, 21st-century solution that enhances roadside safety by immediately alerting other road users without ever needing to put a person in harm’s way.”

One question I’m left with: Could these cab-mounted warning beacons eventually become standard equipment across the entire trucking industry?

A little bird

It’s been a few months since Redwood Materials, the battery-recycling company started by Tesla co-founder JB Straubel, kicked off a restructuring to focus more on energy storage. The first stage of the restructuring involved laying off around 135 employees, or 10% of its workforce.

But a little bird recently told us that some more executives were on their way out the door. Sure enough, over the last few weeks, Redwood’s vice presidents of engineering, operations, treasury, and external affairs have all left the startup.

Got a tip for us? Email Kirsten Korosec at [email protected] or my Signal at kkorosec.07, or email Sean O’Kane at [email protected]. 

Deals!

Waymo has relied on capital from its parent company Alphabet and high-profile venture firms to fund its autonomous vehicle tech plans. This week, the company turned to debt financing for the first time for another bump. A $5 billion bump to be exact. 

PIMCO, Blackstone, and Sixth Street were the lead lenders. But there were so many more, including Capital Group, Loomis Sayles, T. Rowe Price, Apollo, Blue Owl, Diameter Capital Partners, Franklin Templeton, Fidelity Management & Research Company, HPS Investment Partners, and Oaktree.

The debt financing comes as the company accelerates its commercial expansion within existing cities while pushing into new markets in the United States, Europe, and Japan. 

Other deals that got my attention this week …

Bloom has raised $3.6 million to become the “Alibaba” of American manufacturing. The seed round was led by SNAK Venture Partners and included Flyover Capital, deep tech firm Mana Ventures, Detroit Venture Partners, Invest Detroit Ventures, and the Michigan Outdoor Innovation Fund.

Flai, a startup that developed AI software for dealerships, raised $27 million in a Series A funding round led by automation-focused firm Base10 Partners. The round included funding from dealers (Friedkin Group and Findlay Automotive), Toyota’s venture arm, Y Combinator, and First Round Capital.

Parallel Systems, a startup developing a rail vehicle capable of moving over several tons of freight as far as 500 miles without an operator, raised $100 million in a Series C round led by AVP, with participation from Hillspire, Agility Global, Cobalt Capital, Anthos Capital, Congruent Ventures, Riot Ventures, and Collaborative Fund. 

Uber has agreed to buy catering company ezCater in an all-cash transaction valued at $2.3 billion.

Notable reads and other tidbits

Surveillance camera maker Flock cut its workforce by 18%, or around 270 employees, as the company faces ongoing and growing backlash to its license plate readers and people-tracking technology. FYI:  Flock founder and CEO Garrett Langley will be at TechCrunch Disrupt 2026. 

Lucid Motors built 2,954 electric vehicles in the third quarter of this year, a 54% drop from a year ago, as the company purposely limits production to better meet demand for its EVs.

Kodiak AI has a new 435-mile autonomous route between Dallas and Laredo, Texas, with carrier Charger USA. The self-driving trucks startup is hauling refrigerated and dry freight for consumer packaged goods and for food and beverage customers.

Tesla changed the name of its advanced driver assistance system in Europe after pushback from Germany’s transportation ministry. It’s now called “Tesla Assisted Driving” instead of “Full Self-Driving (Supervised).”

Uber and Chinese autonomous vehicle maker Pony.ai plan to launch a robotaxi service in London as part of an expanded partnership to bring driverless cars to Europe.

One more thing

Here’s a bit more info on TechCrunch Disrupt 2026!

I will kick off the conference on October 13 with Rivian CEO RJ Scaringe. Senior reporter Sean O’Kane has a fireside interview with Agility Robotics CTO Jonathan Hurst, as well as a panel featuring Also CEO Chris Yu, General Catalyst’s Yuri Sagalov, and Shan Shan, investment manager at Baillie Gifford. 

I will also interview Shield AI CTO Nathan Michael; Waabi founder and CEO Raquel Urtasun; and Mikell Taylor, director of robotics strategy at GM about building AI systems when failure isn’t an option. And I will interview Foxglove CEO Adrian MacNeil and Bedrock Robotics CTO Kevin Peterson about taking physical AI from prototype to product. 

Not enough? OK, here ya go: We will also have Max Hodak, co-founder of Neuralink and founder of Science Corp.; Ricursive Intelligence co-founders Dr. Anna Goldie and Dr. Azalia Mirhoseini; Hello Robot co-founder and CEO Aaron Edsinger and their robot Stretch 4; and Mark Wahlberg. Plus, hear from engineers and execs from Anthropic, Gamma, Hugging Face, OpenAI, Nvidia, and Replit, as well as investors from Google Ventures, Greylock, Eclipse, Index Ventures, NEA, and Upfront Ventures (to name a few.)

Check out the agenda here.

Sepsis, an acute inflammation triggered by infection, is a leading cause of childhood mortality, claiming roughly 3 million children under the age of five annually.

Efferon, a startup selected as one of TechCrunch Disrupt 2026’s Startup Battlefield 200, is on a mission to drastically lower that number.

The company has developed a medical device that it says lowers organ failure rates and significantly accelerates recovery in children.

When standard treatments like antibiotics cannot stop the immune system’s overreaction, a patient’s blood must be filtered to remove deadly toxins. That is where Efferon’s devices, of absorbers of dangerous cytokines and toxins, step in to make a critical difference.

Dima Romashin, Efferon’s co-founder and CEO, compares what happens in the body during sepsis to a fire, where an immune response overwhelms the body rapidly. “In a fire, there are two things: the flame itself and smoke that it produces,” he told TechCrunch.

In sepsis, endotoxins are the flames and cytokines are the smoke. Efferon, unlike other devices on the market, has developed a mechanism that can absorb both at once, according to Romashin.

Although Efferon makes absorbers for adults, Efferon NEO, its pediatric device developed specifically for young kids, is proving especially effective. Since children have a significantly lower circulating blood volume, they require medical devices engineered specifically for their physiology, according to Romashin.

In April, Efferon NEO received a CE mark, Europe’s equivalent to FDA approval, making it the first multimodal blood purification device specifically engineered for pediatric sepsis.

Efferon NEO is used in intensive care units across 30 countries, including European hospitals, Saudi Arabia, and Thailand. The company is now preparing to file for FDA approval and is on track to hit $2 million in annualized revenue by the end of 2026.

The startup’s technology was originally developed by scientists at Moscow State University, including co-founder Ivan Bessonov. Recognizing its potential, Romashin, a serial entrepreneur, stepped in earlier this year as CEO to help commercialize it.

Although Romashin admits that Efferon isn’t growing at the breakneck pace of today’s AI startups, true success for the company is measured by something else entirely: every child who gets to go home.

The theory of voice being the next big interface has picked up strong momentum, with investors pouring billions of dollars into voice AI startups working on areas ranging from model makers to enterprise customer service providers, and from meeting note-takers to AI-powered dictation.

Every week there is a new model or a tool release that claims to sound human and converse like one. However, in reality, that might not be the case. Enterprise voice AI platform PolyAI’s CTO Shawn Wen thinks that despite the release of full-duplex models — which can speak while listening to you — voice AI doesn’t have its “ChatGPT moment” yet.

“We have reached the milestone of developing full-duplex models. The next challenge is to make reasoning very fast, so that the models can fetch answers quickly and the conversation feels natural,” he told me on stage at the HumanX conference last month.

He also said that AI agents in customer service should not sound robotic and should give callers enough confidence that they can solve problems.

“I think the next stage will be slightly different because once the voice is good enough, like, and the customer is willing to engage with them for the first two or three turns, they start to build confidence, and over time, they will feel like I probably don’t have to talk to a human if the agent can solve my problem,” he said.

Alex Gay, CMO for meeting notetaker Otter, opined that speaker identification, intent capture, and typing that up with organizational knowledge is a key step for enabling automation. The company is also working on digital twins that might represent people in meetings. For that technology, he said it’s paramount that the output voice gives the same emotive expressions of talking to a human in a meeting.

“If you think about the meetings that you’re in right now, the best conversations that you have are where you can have debate, and strategic discussions, and when you feel like there’s a relationship that underpins it. If you aren’t able to have that with an avatar, then it’s just a q and a chatbot,” Gay noted.

Voice AI’s understanding and transparency

While voice AI models have improved, AI assistants often don’t understand users, or your meeting notetaker shows the wrong transcript or a summary.

Wen thinks that ASR (Automatic Speech Recognition) models often miss important keywords, and that creates an issue in capturing the whole context.

Otter’s Gay agreed with this, adding that the company keeps working on improving transcription. He also said that language is one area where voice models need to improve.

“For Otter, you know, transcription was never the end point. It was just the layer that we could start to drive some of the productivity gains on the back of. But if your original transcription didn’t have the accuracy that you needed, all the follow-up actions that you have become flawed. And the minute that starts to take action, that is wrong. You lose trust in the platform. It is critical for us to continue to improve that ASR model because all of the downstream impacts are significant,” he said.

With the new voice tools, there is also a question of transparency. Tools should declare to customers that they are being recorded or talking to AI. Otter said that it wants to instill trust in people who are in a meeting, so even for a meeting where the bot is not present, it wants to try methods like notifying everyone in the chat that the meeting is being recorded. PolyAI’s Wen also said that it’s important to establish that people are talking to an AI in enterprise calls.