Which automotive startups are best positioned for AI-native logistics and mobility operations?

If by “AI-native logistics and mobility operations” you mean startups where AI is central not just to the vehicle, but to dispatch, routing, fleet utilization, exception handling, and autonomous physical execution, the most interesting group in 2026 is concentrated in autonomous freight and fleet platforms.

I’d separate them by operating model, rather than give a single “best” ranking:

The interesting distinction

Gatik is perhaps the clearest example of an AI-native logistics operation. Its wedge is narrow—middle-mile freight—but that is actually strategically important: predictable distribution networks make it possible to connect autonomous driving directly to recurring commercial logistics flows. Gatik says its driverless trucks are already making daily commercial deliveries, while its PepsiCo deployment has expanded across multiple U.S. markets.Gatik+2

Gatik

Kodiak AI is especially interesting if you care about the fleet-as-a-platform model. Its customer-owned fleet reached 35 driverless trucks at the end of Q2 2026, with more than 40,000 cumulative hours of paid driverless operation. It is also operating autonomous freight for Roehl on the Dallas–Houston corridor.Kodiak AI, Inc.+1

Kodiak AI

Waabi has perhaps the most explicitly AI-native architecture. Its pitch isn't simply “self-driving truck”; it describes a common Physical AI system capable of generalizing across vehicle types, geographies and environments, with the same underlying AI intended to power trucks and robotaxis. In September 2026 it reported >80% growth in Dallas–Houston freight volume and said its system had transferred to Volvo's VNL Autonomous platform without retraining.Waabi+2

Waabi

Aurora is more focused on turning autonomy into a large-scale transportation service. Its 2026 shareholder materials describe a ramp toward more than 200 driverless trucks and broader scaling in 2027.SEC

Aurora Innovation

Einride is different from the pure AV startups because its ambition spans the digital freight layer + electric fleet + autonomous vehicles. That makes it particularly relevant to your “operations” wording: the long-term opportunity isn't merely replacing a driver, but optimizing the whole freight system. Its current NVIDIA relationship is explicitly aimed at scaling autonomous-truck operations.The Wall Street Journal

Einride

And Wayve is worth watching from the opposite direction: rather than primarily becoming a carrier, it is increasingly positioning its driving AI as infrastructure that can be licensed to automakers and fleet operators. Recent partnerships include Uber, Nissan and Stellantis.Axios

Wayve

What I would watch as an investor/operator

The key question isn't simply “whose autonomous driving model is best?” It is:

Who owns the feedback loop between AI → vehicle → fleet → logistics network → customer? That creates three potentially powerful models:

1. Autonomous carrier: Gatik, Aurora The startup operates the transportation service and captures operational data.
2. Autonomy-as-a-fleet-platform: Kodiak, Waabi The customer owns/operates vehicles while the startup supplies the AI, hardware and supporting infrastructure.
3. AI-native transportation OS: Einride, with Wayve somewhat adjacent The ambition is to optimize vehicles, routing, utilization, energy and transportation demand as one system.

For AI-native logistics specifically, I'd pay particular attention to Gatik, Kodiak, Waabi and Einride because their propositions extend beyond “make a car drive itself” into recurring freight operations. The evidence is still heterogeneous—miles, driverless hours, contracts, fleet counts and freight volumes aren't directly comparable—so those metrics shouldn't be treated as a common scorecard.note（ノート）+1

If you want, I can also .

Gatik

Kodiak AI is especially interesting if you care about the fleet-as-a-platform model. Its customer-owned fleet reached 35 driverless trucks at the end of Q2 2026, with more than 40,000 cumulative hours of paid driverless operation. It is also operating autonomous freight for Roehl on the Dallas–Houston corridor.Kodiak AI, Inc.+1

Kodiak AI

Waabi has perhaps the most explicitly AI-native architecture. Its pitch isn't simply “self-driving truck”; it describes a common Physical AI system capable of generalizing across vehicle types, geographies and environments, with the same underlying AI intended to power trucks and robotaxis. In September 2026 it reported >80% growth in Dallas–Houston freight volume and said its system had transferred to Volvo's VNL Autonomous platform without retraining.Waabi+2

Waabi

Aurora is more focused on turning autonomy into a large-scale transportation service. Its 2026 shareholder materials describe a ramp toward more than 200 driverless trucks and broader scaling in 2027.SEC

Aurora Innovation

Einride is different from the pure AV startups because its ambition spans the digital freight layer + electric fleet + autonomous vehicles. That makes it particularly relevant to your “operations” wording: the long-term opportunity isn't merely replacing a driver, but optimizing the whole freight system. Its current NVIDIA relationship is explicitly aimed at scaling autonomous-truck operations.The Wall Street Journal

Einride

And Wayve is worth watching from the opposite direction: rather than primarily becoming a carrier, it is increasingly positioning its driving AI as infrastructure that can be licensed to automakers and fleet operators. Recent partnerships include Uber, Nissan and Stellantis.axios.com

Wayve

What I would watch as an investor/operator

The key question isn't simply “whose autonomous driving model is best?” It is:

Who owns the feedback loop between AI → vehicle → fleet → logistics network → customer? That creates three potentially powerful models:

If by “AI-native logistics and mobility operations” you mean companies where AI is becoming the operating layer—not merely an ADAS feature—the strongest current set is concentrated in autonomous freight and software-defined fleet orchestration.

I’d look at these companies:

The interesting distinction

Kodiak is particularly notable because it has moved beyond demonstrations into measurable driverless commercial operations. As of Q2 2026, it reported 35 customer-owned driverless vehicles, more than 40,000 hours of paid driverless operation and more than 20,000 loads delivered. Its Atlas deployment is also scaling toward 100 autonomous trucks.Kodiak AI, Inc.+1

Waabi is interesting from a pure-AI architecture perspective. Its September 2026 update says freight volumes on its Dallas–Houston operation had grown more than 80% during 2026, while its system demonstrated zero-shot transfer to Volvo's VNL Autonomous platform without retraining. That kind of vehicle/platform generalization is potentially important for becoming an operating layer rather than a single-vehicle product.Waabi

Einride is arguably the clearest example of the broader logistics operating system thesis. Its Saga AI platform coordinates electric fleets, charging, energy and freight operations rather than treating autonomy as the entire product. By mid-2026, it reported more than 5,400 driverless operating hours and 18.5 million planned electric miles; it subsequently announced plans for 500 Tesla Semis on Saga AI.Einride+1

Gatik occupies a different but important position: rather than attacking the entire long-haul network, its focus is repeatable middle-mile routes where autonomous vehicles can be tightly integrated into distribution operations. That's an attractive environment for an AI-native logistics system because routes, facilities, loads and operating conditions are relatively structured.

What I would actually investigate

Rather than treating “autonomous vehicle startup” as the category, I'd evaluate companies against five operational capabilities:

1. Autonomy — Can the system actually remove the driver, rather than merely assist one?
2. Fleet intelligence — Does AI optimize dispatch, routing, charging, maintenance and utilization?
3. Network integration — Is it embedded in a shipper's actual logistics workflow?
4. Data flywheel — Do real-world operations continuously improve the system?
5. Asset-light scalability — Can the software orchestrate vehicles from multiple OEMs rather than requiring the company to own every truck?

On that framework, Einride is the clearest “AI logistics operating system” case, while Kodiak and Waabi are especially interesting for the autonomous-driving/physical-AI layer. Gatik is particularly relevant if your thesis is structured middle-mile automation rather than long-haul autonomy.

One important caveat: these aren't all at the same stage. Some have substantial commercial driverless operations, while others are still moving from pilots toward scaled deployment; the autonomous-trucking sector is only now entering broader commercialization.TT News

If you're evaluating this from an investment/venture perspective, I can also break these companies down by technical moat, customer traction, capital intensity, revenue model, and likelihood of becoming the “operating system” for autonomous fleets—without ranking them.

Einride is arguably the clearest example of the broader logistics operating system thesis. Its Saga AI platform coordinates electric fleets, charging, energy and freight operations rather than treating autonomy as the entire product. By mid-2026, it reported more than 5,400 driverless operating hours and 18.5 million planned electric miles; it subsequently announced plans for 500 Tesla Semis on Saga AI.

The automotive and mobility startups best positioned for AI-native logistics and operations are those combining physical autonomy, real-time agentic decision intelligence, and unified fleet data platforms. Rather than relying on legacy retrofits or passive telematics, these companies build systems where foundational machine learning and physical AI drive automated execution from the ground up.

Top Startups in AI-Native Logistics and Mobility

• Nuro : Builds universal, vehicle-agnostic autonomous driving systems designed to license out to automakers and commercial fleets. By focusing purely on scaled robotic delivery and low-occupant transport, Nuro avoids passenger-car complexities to optimize straight-to-destination AI logistics.tracxn.com
• Proaction : Operates as an AI-native fleet management platform built to replace legacy, fragmented software stacks for commercial operators. It provides consolidated operational visibility and automated cost-reduction analytics tailored to modern commercial fleets.projectstartups.com
• Arqh : Delivers an AI decision intelligence platform for logistics teams that replaces manual phone calls and spreadsheets with a hybrid agentic optimization system. It uses large language models to orchestrate real-time freight and routing decisions dynamically.ventureradar.com
• Self Inspection : Powers an AI-powered vehicle inspection service leveraging standard smartphone cameras to autonomously detect, log, and quantify vehicle and fleet damage. This automates the critical intake and turnaround bottleneck in automotive rental, logistics, and remarketing workflows.projectstartups.com
• Axe : Provides an AI voice assistant built specifically to automate dispatch calls and coordinate live delivery tracking through natural, conversational speech. It removes human friction points in real-time driver-to-hub communication chains.ventureradar.com

Core Capabilities Comparison

Would you like to narrow this focus down by geographic region (e.g., North America vs. Europe) or a specific vertical like long-haul trucking versus last-mile delivery?

The automotive and mobility landscape has shifted from hardware-first experimentation to AI-native autonomy and software-defined logistics . The best-positioned players are those leveraging end-to-end deep learning, simulation-first training, and scalable Driver-as-a-Service (DaaS) frameworks.landbase.com+2

The automotive startups and scale-ups leading the charge in AI-native logistics and mobility operations include:

1. Wayve
   • Focus: Embodied AI and foundation models for autonomous driving.
   • Why they are positioned well: Unlike traditional rule-based autonomy stack providers, Wayve utilizes end-to-end deep learning (similar to LLMs, but for driving physics and decision-making). Their approach allows generalizable scaling across multiple vehicle platforms and international cities without needing hyper-specific HD maps for every lane.landbase.com+2
2. Aurora
   • Focus: Long-haul autonomous trucking and commercial mobility.
   • Why they are positioned well: Powered by their proprietary "Aurora Driver" system, they targeted the B2B logistics sector where unit economics and driver shortages make automated line-hauling immediately viable. Their strategic consolidation (including acquiring Uber’s advanced self-driving division) positions them firmly at the forefront of commercial freight integration.greyb.com+2
3. Waabi
   • Focus: Simulation-first autonomous driving.
   • Why they are positioned well: Waabi bypasses the physical limits of road-testing miles by utilizing generative AI and a "world simulator" (Waabi World) to train software safely on edge cases, rare hazards, and complex physics before hitting public asphalt. This generative-AI-native training loop accelerates deployment across various vehicle categories drastically faster than legacy competitors.landbase.com+1
4. Gatik
   • Focus: B2B middle-mile logistics and autonomous freight.
   • Why they are positioned well: Gatik specializes in high-frequency, predictable B2B regional routes (such as hub-to-store distribution for retail giants). By zeroing in on fixed B2B logistics corridors rather than unpredictable consumer robotaxis, they have achieved robust commercialization and operational unit profitability.landbase.com+1
5. Kodiak AI
   • Focus: Modular autonomous trucking (Driver-as-a-Service).
   • Why they are positioned well: Kodiak focuses on a hardware-agnostic, lightweight "Kodiak Driver" model that can be integrated into existing fleet trucks. Their dual focus on commercial freight and defense applications proves the resilience and adaptability of their AI stack in chaotic environments.landbase.com+1
6. Applied Intuition
   • Focus: Autonomy development infrastructure and physical AI platforms.
   • Why they are positioned well: While not an operator of fleets themselves, they provide the essential AI-native simulation, validation, and safety-assurance tooling that powers nearly every next-gen automotive and logistics workflow. They function as the picks-and-shovels backbone for the entire ecosystem.landbase.com+1

If you'd like to dive deeper, let me know:

• Are you looking at this from an investment/market analysis perspective or a technical supply chain angle?
• Would you like to focus more on passenger mobility (robotaxis) or freight/trucking logistics?

If by “AI-native logistics and mobility operations” you mean companies that can turn autonomous vehicles into a managed operating system for moving freight or people—not merely build an impressive self-driving stack—my ranking as of September 2026 is:

1. Gatik — best positioned for an AI-native logistics operator

Gatik is my top pick if the thesis is “autonomy becomes a logistics operating business.”

The important distinction is that Gatik isn't just selling autonomy technology. It operates freight networks: repetitive distribution-center-to-store and regional routes where the value of autonomy can be measured in delivery reliability, vehicle utilization, capacity and cost.

Its position has strengthened dramatically in 2026. Gatik says it has more than $600 million in contracted revenue, 85,000 fully driverless orders and 99% on-time delivery, while PepsiCo signed a multiyear agreement covering autonomous freight across its North American supply chain.Gatik+1

It also raised another $200 million in August 2026, giving it capital to expand beyond its current markets.TechCrunch+1

Why I like it: Gatik has chosen a narrower operational problem where autonomy is economically useful before generalized robotaxis or fully autonomous long-haul trucking are solved.

---

2. Aurora — best shot at the autonomous freight network

Aurora Innovation has the strongest case if you're betting on nationwide autonomous freight infrastructure.

Aurora has moved into commercial scaling with its second-generation driverless trucks and says it expects to exit 2026 with 200 driverless trucks. Its Roush manufacturing operation is targeting a 1,000-truck annual run rate, while Volvo Autonomous Solutions plans driverless Volvo VNL operations powered by Aurora beginning in Q1 2027.Aurora Innovation, Inc.+1

More importantly for the “AI-native operations” thesis, Aurora is working beyond the highway-driving problem: dynamic rerouting, customer-facility operations, fueling, weigh stations and increasingly end-to-end freight workflows.Aurora Innovation, Inc.+1

The strategic advantage is asset-light scalability: Aurora intends its Driver to operate across multiple truck OEM platforms rather than being locked to one vehicle.Aurora Innovation, Inc.

The risk: Gatik is ahead in the boring-but-crucial business of repeatedly moving commercial freight without a driver. Aurora still has to prove that its enormous long-haul opportunity can translate into durable unit economics.

---

3. Kodiak AI — the sleeper with unusually deep operational data

Kodiak AI may have the most interesting physical-AI-to-operations feedback loop.

By Q2 2026, Kodiak had 35 customer-owned driverless trucks, more than 40,000 hours of paid driverless operation, over 20,000 loads and more than 300,000 tons of freight delivered in that quarter alone.Kodiak AI, Inc.

That's valuable because autonomous logistics isn't ultimately won by the prettiest demo. It's won through the accumulation of operational edge cases:

route → load → drive → exception → intervention → data → simulation → model improvement → lower intervention rate → more utilization.

Kodiak's BreakPoint simulation system, for example, can run more than a million simulations per hour, according to the company.Kodiak AI, Inc.

Its expansion into long-haul, Canada, logging, oilfield logistics and defense also gives it multiple environments in which to reuse the same autonomy stack.Kodiak AI+1

Why it isn't #1: Gatik has a more compelling commercial network story, while Aurora has greater potential scale.

---

4. Applied Intuition — potentially the biggest picks-and-shovels winner

This is the company I'd watch if your definition of “automotive startup” includes vehicle AI infrastructure.

Applied Intuition isn't primarily trying to operate autonomous fleets. It is building the software infrastructure that lets automakers and autonomous-vehicle companies develop, simulate, validate and deploy intelligent vehicles.

Its 2025 figures are striking: customers conducted more than 50 million simulations, while its platforms handled hundreds of petabytes of training data and trillions of requests.Applied Intuition

It also already serves 18 of the top 20 automotive OEMs, giving it unusually broad distribution into the eventual autonomous-vehicle ecosystem.Applied Intuition

Its strategic position is therefore different:

Gatik/Aurora/Kodiak = AI-native logistics operators Applied Intuition = AI-native infrastructure layer for the vehicles If autonomy becomes ubiquitous, the second position could be extraordinarily valuable.

---

5. Torc Robotics — particularly interesting because of OEM integration

Torc Robotics has an advantage that pure-play startups often lack: it is an independent subsidiary of Daimler Truck.

Torc is concentrating on the Freightliner Cascadia and developing a production-oriented autonomous system rather than treating the truck as a research platform. Its stack combines autonomous-driving software, data loops and generative-AI simulation.Torc Robotics+1

That makes Torc particularly attractive if you believe the winning model is:

OEM + autonomy company + fleet operators + standardized production vehicle.

Its weakness relative to Aurora/Gatik/Kodiak is simply that its commercial deployment is not yet as visible.

---

6. Waabi — potentially the highest-upside AI bet

Waabi is worth separating from the pack because its thesis is especially AI-native rather than robotics-first.

Its approach centers on AI simulation and learning rather than simply accumulating enormous quantities of real-world driving miles. If that approach produces comparable safety and reliability with dramatically less real-world data and testing, the economics could be powerful.

But I'd currently assign it more technology optionality than proven operating leverage. Gatik, Aurora and Kodiak have more compelling evidence that their systems are becoming businesses rather than engineering programs.

---

The real dividing line

I would actually group the market into three tiers:

A. AI-native logistics operators

• Gatik
• Aurora
• Kodiak

These have the best chance of owning the customer relationship + vehicle + autonomy + dispatch/operations loop.

B. Autonomous vehicle platforms

• Torc
• Waabi
• Stack AV

Potentially excellent technology, but the ultimate economics depend more heavily on OEM/fleet relationships.

C. AI infrastructure

• Applied Intuition

Less exposed to whether any one autonomous fleet wins. It can potentially sell the underlying development/simulation/vehicle-intelligence infrastructure to many winners.

My investment-style ranking

If I were evaluating them specifically around the thesis “AI will reorganize logistics and mobility operations,” rather than simply “who has the best self-driving technology,” I'd score them:

Gatik > Aurora > Kodiak > Applied Intuition > Torc > Waabi > Stack AV

The interesting thing is that Gatik doesn't necessarily need to solve generalized autonomy to win. It needs to make a finite set of logistics networks dramatically cheaper and more reliable. That is a much more tractable path to an AI-native operating company.

Aurora has the bigger potential market and arguably the strongest path to a national freight network, while Kodiak has one of the strongest real-world operational feedback loops.

And I'd keep Applied Intuition in a separate category: it could be the most valuable company on this list even if it never operates a single truck, because it potentially supplies the software infrastructure for the entire AI-defined vehicle industry.Applied Intuition+1

Bottom line: Gatik is the strongest current “AI-native logistics operator”; Aurora is the strongest scale bet; Kodiak is the strongest operational dark horse; Applied Intuition is the strongest infrastructure bet.

---

3. Kodiak AI — the sleeper with unusually deep operational data

Kodiak AI may have the most interesting physical-AI-to-operations feedback loop.

By Q2 2026, Kodiak had 35 customer-owned driverless trucks, more than 40,000 hours of paid driverless operation, over 20,000 loads and more than 300,000 tons of freight delivered in that quarter alone.Kodiak AI, Inc.

That's valuable because autonomous logistics isn't ultimately won by the prettiest demo. It's won through the accumulation of operational edge cases:

route → load → drive → exception → intervention → data → simulation → model improvement → lower intervention rate → more utilization.

Kodiak's BreakPoint simulation system, for example, can run more than a million simulations per hour, according to the company.Kodiak AI, Inc.

Its expansion into long-haul, Canada, logging, oilfield logistics and defense also gives it multiple environments in which to reuse the same autonomy stack.Kodiak AI+1

Why it isn't #1: Gatik has a more compelling commercial network story, while Aurora has greater potential scale.

---

4. Applied Intuition — potentially the biggest picks-and-shovels winner

This is the company I'd watch if your definition of “automotive startup” includes vehicle AI infrastructure.

Applied Intuition isn't primarily trying to operate autonomous fleets. It is building the software infrastructure that lets automakers and autonomous-vehicle companies develop, simulate, validate and deploy intelligent vehicles.

Its 2025 figures are striking: customers conducted more than 50 million simulations, while its platforms handled hundreds of petabytes of training data and trillions of requests.Applied Intuition

It also already serves 18 of the top 20 automotive OEMs, giving it unusually broad distribution into the eventual autonomous-vehicle ecosystem.Applied Intuition

Its strategic position is therefore different:

Gatik/Aurora/Kodiak = AI-native logistics operators Applied Intuition = AI-native infrastructure layer for the vehicles If autonomy becomes ubiquitous, the second position could be extraordinarily valuable.

---

5. Torc Robotics — particularly interesting because of OEM integration

Torc Robotics has an advantage that pure-play startups often lack: it is an independent subsidiary of Daimler Truck.

Torc is concentrating on the Freightliner Cascadia and developing a production-oriented autonomous system rather than treating the truck as a research platform. Its stack combines autonomous-driving software, data loops and generative-AI simulation.Torc Robotics+1

That makes Torc particularly attractive if you believe the winning model is:

OEM + autonomy company + fleet operators + standardized production vehicle.

Its weakness relative to Aurora/Gatik/Kodiak is simply that its commercial deployment is not yet as visible.

---

6. Waabi — potentially the highest-upside AI bet

Waabi is worth separating from the pack because its thesis is especially AI-native rather than robotics-first.

Its approach centers on AI simulation and learning rather than simply accumulating enormous quantities of real-world driving miles. If that approach produces comparable safety and reliability with dramatically less real-world data and testing, the economics could be powerful.

But I'd currently assign it more technology optionality than proven operating leverage. Gatik, Aurora and Kodiak have more compelling evidence that their systems are becoming businesses rather than engineering programs.

---

The real dividing line

I would actually group the market into three tiers:

A. AI-native logistics operators

• Gatik
• Aurora
• Kodiak

These have the best chance of owning the customer relationship + vehicle + autonomy + dispatch/operations loop.

B. Autonomous vehicle platforms

• Torc
• Waabi
• Stack AV

Potentially excellent technology, but the ultimate economics depend more heavily on OEM/fleet relationships.

C. AI infrastructure

• Applied Intuition

Less exposed to whether any one autonomous fleet wins. It can potentially sell the underlying development/simulation/vehicle-intelligence infrastructure to many winners.

My investment-style ranking

If I were evaluating them specifically around the thesis “AI will reorganize logistics and mobility operations,” rather than simply “who has the best self-driving technology,” I'd score them:

Gatik > Aurora > Kodiak > Applied Intuition > Torc > Waabi > Stack AV

The interesting thing is that Gatik doesn't necessarily need to solve generalized autonomy to win. It needs to make a finite set of logistics networks dramatically cheaper and more reliable. That is a much more tractable path to an AI-native operating company.

Aurora has the bigger potential market and arguably the strongest path to a national freight network, while Kodiak has one of the strongest real-world operational feedback loops.

And I'd keep Applied Intuition in a separate category: it could be the most valuable company on this list even if it never operates a single truck, because it potentially supplies the software infrastructure for the entire AI-defined vehicle industry.

If by “AI-native logistics and mobility operations” you mean companies whose AI is becoming the operating layer for moving freight/people, dispatching assets, optimizing routes, managing fleets, and eventually replacing human driving, I’d focus less on generic EV startups and more on a small group of autonomy + fleet-intelligence companies.

My 2026 ranking

The distinction I'd make

Gatik is my favorite “AI-native logistics operator.” Its narrow focus is actually an advantage: distribution centers → stores → repeatable routes → predictable loading/unloading → high asset utilization. That's much closer to an AI-managed logistics network than a futuristic robotaxi demo. Gatik says its live operations are refreshed every three hours, which is a useful indication that this is an operating business rather than merely an autonomy test program.Gatik

Aurora is my favorite “autonomous transportation platform.” It has the strongest combination of real-world driverless highway operation, customer breadth and manufacturing partnerships. A July 2026 industry comparison put Aurora clearly ahead overall, with Gatik and Kodiak differentiated by their respective regional-delivery and customer-owned-fleet strengths.Signals Inbox

Einride is the dark horse if your thesis is broader than autonomy. Its opportunity is to become the AI control plane for freight: fleet selection, routing, charging/energy, utilization and eventually autonomous vehicles. Its August 2026 announcement of 500 Tesla Semis on its Saga AI platform is particularly important because it expands the asset base that its software can orchestrate.Einride

What I'd actually bet on

If I were evaluating these as platform businesses, rather than simply asking “who has the best self-driving technology?”, I'd use this framework:

1. Gatik — best near-term AI logistics operator
2. Einride — best AI freight orchestration thesis
3. Aurora — best autonomy technology + scale combination
4. Kodiak — best fleet-integrated autonomy
5. Waabi — highest technical optionality
6. PlusAI — strongest OEM-distribution option
7. Stack AV — high upside, insufficient public evidence

The really valuable company may ultimately be the one that owns the decision layer above the vehicle. Driving is only one optimization problem. The bigger opportunity is deciding which vehicle should move which load, when, along what route, using what energy, with what human intervention, and how the entire fleet should respond to changing demand.

That's why I'd watch Gatik and Einride especially closely: their strategic trajectories point toward AI managing a logistics network, rather than AI merely controlling a car or truck.

Aurora is my favorite “autonomous transportation platform.” It has the strongest combination of real-world driverless highway operation, customer breadth and manufacturing partnerships. A July 2026 industry comparison put Aurora clearly ahead overall, with Gatik and Kodiak differentiated by their respective regional-delivery and customer-owned-fleet strengths.signalsinbox.com

Einride is the dark horse if your thesis is broader than autonomy. Its opportunity is to become the AI control plane for freight: fleet selection, routing, charging/energy, utilization and eventually autonomous vehicles. Its August 2026 announcement of 500 Tesla Semis on its Saga AI platform is particularly important because it expands the asset base that its software can orchestrate.

What I'd actually bet on

If I were evaluating these as platform businesses, rather than simply asking “who has the best self-driving technology?”, I'd use this framework:

If by “AI-native logistics and mobility operations” you mean companies that can use AI not merely to make a car drive itself, but to run the transportation system—dispatching vehicles, optimizing routes, managing charging, predicting maintenance, coordinating fleets, and continuously learning from operations—my ranking in September 2026 would be:

1. Einride — strongest operations thesis

Einride is unusually close to the thing I'd call an AI-native logistics operating system rather than merely an autonomous-vehicle company. Its Saga platform connects shipments, vehicles, charging infrastructure and autonomous trucks, and continuously plans, monitors and optimizes freight operations.Einride+1

The recent launch of Flip AI, an agentic layer for fleet, shipper and charging workflows, is particularly important: it suggests Einride is moving from “software that optimizes a fleet” toward AI agents that actually operate the fleet.Einride

That's the architecture I think ultimately wins: AI driver + fleet OS + energy + dispatch + logistics planning.

2. Gatik — strongest proof that the model works

Gatik has perhaps the clearest evidence that AI-native logistics can become a real operating business. It is running driverless freight networks for major retailers and grocers in the U.S. and Canada, and raised a $200M Series D in August 2026.Gatik

Its focus is also strategically smart: repeatable middle-mile routes, rather than trying to solve every possible driving environment simultaneously.

If you're looking for the startup most likely to demonstrate the economics of autonomous logistics before universal robotaxis arrive, I'd put Gatik near the top.

3. Applied Intuition — potentially the most important infrastructure company

Applied Intuition is a different bet. It isn't primarily trying to own the fleet; it's building the software layer used to develop and operate autonomous vehicles and other physical machines.

That's powerful because the winner of AI-native transportation doesn't necessarily need to own trucks or robotaxis. The higher-margin position could be the infrastructure through which automakers, autonomous fleets and industrial operators build their systems.

I'd therefore distinguish:

• Best transportation operator: Einride/Gatik
• Best underlying software platform: Applied Intuition

4. Wayve — strongest general-purpose mobility AI

wayve.ai is arguably the most interesting pure embodied-AI play. Its approach is vehicle-agnostic, mapless and designed to generalize across cities and vehicle types. It has also raised $1.5B in 2026 and is moving into commercial robotaxi deployment with Uber.Wayve

The important thing isn't just autonomous driving. If Wayve's model generalizes as advertised, the same intelligence could eventually power:

vehicle → fleet → mobility network → logistics network.

That's an enormous TAM, although its logistics-operations footprint is less mature than Einride's or Gatik's.

5. Waabi — biggest technical wildcard

waabi.ai has one of the most explicitly AI-native architectures: a single reasoning-oriented model intended to generalize across trucks and robotaxis, combined with neural simulation.Waabi

I like the thesis enormously, but I'd give it a lower operational score than Gatik or Einride because commercial deployment at sustained scale is ultimately the proof point.

6–8. Nuro, PlusAI and Kodiak

Nuro has a particularly attractive licensing strategy. Rather than becoming another capital-intensive robotaxi operator, it is positioning the Nuro Driver as a vehicle-agnostic autonomy platform that automakers and mobility companies can license.Nuro

PlusAI is compelling on the trucking side because it combines an AI-native driving stack with major OEM relationships. It reports 7M+ autonomy miles and partnerships spanning TRATON/Scania/MAN/International, Hyundai and IVECO. PlusAI Its September 2026 SPAC announcement also puts it on a different commercialization trajectory.Reuters

Kodiak has demonstrated actual paid driverless operations and is pursuing a turnkey autonomy model for fleets. Its 2025 results included 10 driverless trucks, 5,200+ hours of paid driverless service and 3M+ autonomous miles.Kodiak AI

The bigger picture

I'd actually divide the market into three different winning strategies:

• AI logistics operator: Einride, Gatik
• AI autonomy platform: Wayve, Waabi, Nuro, PlusAI, Kodiak
• AI infrastructure/platform: Applied Intuition

And if the question is “Who could become the AWS/Palantir of physical transportation?”, rather than “Who will build the best self-driving truck?”, my shortlist changes to:

Applied Intuition → Einride → Wayve → Nuro → Waabi.

The key metric I'd watch isn't autonomous miles. It's AI-managed vehicle-hours per dollar of human operational labor. The company that can autonomously make the decisions around where vehicles go, when they charge, which jobs they accept, how they are maintained, and how the network responds to exceptions has a much larger opportunity than a company that merely replaces the driver.

I like the thesis enormously, but I'd give it a lower operational score than Gatik or Einride because commercial deployment at sustained scale is ultimately the proof point.

6–8. Nuro, PlusAI and Kodiak

Nuro has a particularly attractive licensing strategy. Rather than becoming another capital-intensive robotaxi operator, it is positioning the Nuro Driver as a vehicle-agnostic autonomy platform that automakers and mobility companies can license.Nuro

PlusAI is compelling on the trucking side because it combines an AI-native driving stack with major OEM relationships. It reports 7M+ autonomy miles and partnerships spanning TRATON/Scania/MAN/International, Hyundai and IVECO. plus.ai Its September 2026 SPAC announcement also puts it on a different commercialization trajectory.reuters.com

Kodiak has demonstrated actual paid driverless operations and is pursuing a turnkey autonomy model for fleets. Its 2025 results included 10 driverless trucks, 5,200+ hours of paid driverless service and 3M+ autonomous miles.

The bigger picture

I'd actually divide the market into three different winning strategies: