DRIVERLESS TRUCKS ARE HERE: Gatik Raises $200 Million to Scale AI-Powered Freight

 

The race to automate transportation is entering a new stage.

Autonomous trucking company Gatik has raised $200 million in Series D funding to dramatically expand its fleet of fully driverless delivery trucks across North America.

The funding round was led by the Qatar Investment Authority and Koch Disruptive Technologies, with participation from Millennium Management, ARK Invest, Intact Private Capital, and other investors.

But the biggest story is not the investment.

It is that these trucks are already working.

Gatik says it has completed approximately 85,000 fully driverless orders, generated more than $600 million in contracted revenue, and achieved a 99 percent on-time delivery rate across its operations. These performance figures come from the company itself.

Gatik currently operates dozens of driverless trucks across the United States and Canada.

The company told Reuters that it is targeting more than 100 fully driverless trucks by the end of 2026, with plans to eventually expand its fleet into the thousands.

This is no longer science fiction.

AI is beginning to drive commercial freight without a human sitting behind the wheel.

Not Robotaxis. These AI Vehicles Carry Your Groceries

Most public discussions about autonomous vehicles focus on robotaxis.

Waymo.

Tesla.

Zoox.

But Gatik is attacking a different part of the transportation industry.

Freight.

More specifically, what the logistics industry calls the middle mile.

The middle mile is the movement of products between facilities.

For example:

Warehouse → Distribution Center → Store

rather than:

Store → Customer's House

That distinction is important.

Middle-mile freight often involves repetitive, high-frequency routes between known commercial locations.

This makes it particularly attractive for autonomous driving.

The truck does not necessarily need to understand every road in America.

It needs to reliably understand and operate within specific environments where the system has been thoroughly tested.

These Trucks Have No Driver or Safety Observer

Gatik's Level 4 autonomous trucks can operate without a human driver or safety observer onboard when they are inside their approved operating conditions.

This is an important detail.

Level 4 does not mean the truck can drive anywhere, anytime, under every imaginable condition.

Autonomous vehicles operate within what engineers call an Operational Design Domain, or ODD.

That defines the conditions where the autonomous system is allowed to operate.

Those conditions may include:

• specific roads
• geographic locations
• weather conditions
• traffic environments
• operating speeds
• defined routes or networks

Inside that validated environment, the AI performs the driving task.

The company's current third-generation trucks can operate across both highways and surface streets, according to Gatik CEO Gautam Narang, including in some light rain and snow conditions.

From 10-Mile Routes to 400-Mile Networks

What makes Gatik's development particularly interesting is how far the technology has evolved.

Its early autonomous delivery systems operated on short and predictable routes.

In 2020, Gatik and Canadian retailer Loblaw launched five vehicles operating across five predetermined routes with fixed pickup and delivery locations.

The trucks initially carried safety drivers.

By 2022, Gatik and Loblaw had moved into fully driverless commercial operations transporting ambient, refrigerated, and frozen products between a distribution center and nearby stores.

Now the network is becoming considerably more sophisticated.

TechCrunch reports that Gatik has evolved from fixed trips shorter than 10 miles into dynamic networks with dozens of pickup and delivery points covering distances of up to 400 miles.

That is a major transition.

The AI is no longer simply memorizing one repetitive route.

The logistics system can increasingly respond to changing commercial requirements.

PepsiCo Is Already Using Driverless Freight

One of Gatik's most important customers is PepsiCo.

In June 2026, the companies announced a multi-year commercial agreement to deploy driverless freight vehicles across PepsiCo's North American supply chain.

According to TechCrunch, the operation includes 41 fully driverless box trucks transporting Frito-Lay products between distribution centers and stores in Dallas, Phoenix, and Northwest Arkansas.

Think about what this means.

Products such as chips and other consumer goods can leave a distribution facility inside a truck with:

No human driver onboard.

The AI drives.

The truck transports the cargo.

The vehicle reaches the next facility.

The supply chain continues.

And this is happening commercially, not merely on a closed testing track.

Canada Is Scaling It Too

Gatik has also significantly expanded its partnership with Canada's largest retailer, Loblaw.

In September 2025, the companies announced a five-year expansion agreement involving an initial deployment of 50 autonomous trucks across the Greater Toronto Area.

Twenty were scheduled for deployment by the end of 2025, followed by another 30 by the end of 2026.

The expanded network is intended to support more than 300 Loblaw stores.

The companies began autonomous deliveries together in 2020.

Before moving into fully driverless operations, they reported completing more than 150,000 autonomous deliveries with safety drivers onboard.

The evolution is clear:

Safety driver → Autonomous operation → Driverless operation → Commercial scaling

That progression is what separates today's autonomous freight industry from many earlier self-driving demonstrations.

Artificial Intelligence Is the Driver

At the center of Gatik's trucks is its proprietary autonomous driving technology called Gatik Driver.

The system processes data from the truck's sensors and makes decisions about:

• lanes
• vehicles
• pedestrians
• intersections
• traffic signals
• obstacles
• road conditions
• speed
• braking
• acceleration
• navigation

The goal is to create an AI system capable of safely performing the complete driving task within its approved operating environment.

That requires more than recognizing objects.

The AI must continuously answer questions such as:

What is happening around me?

What will other road users probably do next?

What action should I take?

Is it safe to continue?

All of this happens while the truck is moving.

NVIDIA Is Supplying Some of the AI Computing Power

Gatik is also working closely with NVIDIA.

The company's autonomous Class 6 and Class 7 trucks are integrating NVIDIA DRIVE AGX for onboard AI processing.

The platform processes the enormous streams of information produced by cameras and other sensors while running autonomous driving workloads in real time.

That relationship illustrates another major AI trend.

NVIDIA's chips are no longer only powering ChatGPT-style systems inside data centers.

Its computing platforms are increasingly being placed inside:

Cars. Trucks. Robots. Factories. Autonomous machines.

Generative AI may have made NVIDIA famous to the wider public.

Physical AI could dramatically expand where NVIDIA's technology operates.

Gatik Is Training AI Inside Virtual Worlds

Driving millions of kilometers in the real world is expensive.

It is also impossible to guarantee that a vehicle will encounter every dangerous situation during testing.

What happens if a pedestrian suddenly runs across the road?

What about heavy fog?

Snow?

A cyclist weaving through traffic?

A broken traffic signal?

Construction?

Sensor failure?

These events may be rare, but autonomous systems must still know how to respond.

So Gatik built a simulation system called Arena.

Launched in July 2025, Arena allows engineers to generate realistic synthetic driving environments for training and validating autonomous driving technology.

The platform integrates NVIDIA Cosmos world foundation models to create sophisticated virtual environments.

Engineers can simulate scenarios involving:

• rain
• fog
• snow
• poor visibility
• difficult intersections
• pedestrians
• cyclists
• animals
• construction zones
• emergency vehicles
• unusual road behavior
• degraded sensors

And unlike an actual road event, a simulated scenario can be repeated thousands of times.

AI Can Practice Crashing Without Crashing

This is one of the most powerful ideas behind simulation.

Suppose engineers want to test what happens when:

A truck approaches an intersection.

It is raining.

Visibility is poor.

A pedestrian suddenly enters the road.

Another vehicle changes lanes.

One sensor provides degraded information.

Testing this situation repeatedly using actual trucks would be dangerous and expensive.

Inside simulation, engineers can recreate it again and again.

Change the speed.

Change the rain.

Change the pedestrian position.

Change traffic.

Change sensor conditions.

Then observe how the autonomous system responds.

Synthetic data is therefore becoming an important component of Physical AI development.

Isuzu Wants to Mass-Produce Autonomous Trucks

Gatik is also working with Japanese commercial vehicle manufacturer Isuzu Motors.

Isuzu invested $30 million in Gatik in 2024 as part of their collaboration on autonomous logistics and Level 4 commercial vehicles in North America.

The partnership is significant because software alone cannot make a truly driverless truck safe.

The truck itself needs redundancy.

If one critical component fails, another system must be able to respond safely.

Steering.

Brakes.

Power.

Computers.

Sensors.

Autonomous commercial vehicles therefore require hardware architectures designed specifically around machines rather than human drivers.

This represents another transition.

We are moving from:

Putting autonomous software into ordinary trucks

toward:

Building trucks specifically designed to be driven by AI.

Driverless Trucks Could Work 24/7

There is also an obvious economic reason companies are interested.

Human drivers need sleep.

They need breaks.

They work shifts.

They get sick.

They have legal limits on how long they can drive.

An autonomous vehicle potentially changes the economics of freight transportation.

A properly maintained driverless truck could theoretically operate for far longer periods, limited primarily by:

• loading and unloading
• maintenance
• fueling or charging
• road regulations
• environmental conditions
• its operational design domain

Gatik says its third-generation vehicles can operate around the clock within their supported environments.

For logistics companies, that could dramatically increase vehicle utilization.

But What Happens to Truck Drivers?

This is where the story becomes much more complicated.

If autonomous freight scales from dozens of vehicles to thousands, and eventually tens of thousands, some driving jobs could clearly face pressure.

But automation rarely eliminates an entire industry overnight.

It changes the structure of work.

Autonomous fleets still require people for:

• maintenance
• fleet operations
• remote assistance
• safety engineering
• logistics management
• cybersecurity
• AI development
• mapping
• sensor maintenance
• warehouse operations
• regulatory compliance

The key question is not simply:

"Will AI eliminate jobs?"

The more important question is:

"Which tasks will AI perform, and what new skills will humans need when it does?"

That same question is already affecting office workers through generative AI.

Physical AI will bring it to transportation, manufacturing, warehouses, agriculture, and logistics.

Why the Philippines Should Pay Attention

The Philippines should watch autonomous freight carefully.

Our economy depends heavily on moving goods.

Every day, enormous volumes of products move between:

ports, warehouses, factories, distribution centers, supermarkets, malls, airports, industrial parks, and provincial hubs.

Imagine autonomous freight operating between:

Port of Manila → Distribution Center

Clark → Subic

Batangas Port → CALABARZON industrial zones

Warehouse → Supermarket

Factory → Logistics hub

This does not mean driverless trucks are ready to suddenly operate everywhere in the Philippines.

Our roads present challenges very different from controlled North American freight networks.

Traffic behavior.

Road quality.

Motorcycles.

Pedestrians.

Weather.

Flooding.

Informal road activity.

Infrastructure.

Regulation.

Mapping.

These make autonomous driving far more difficult.

But that is precisely why Filipino engineers, universities, transportation planners, and AI developers should be studying the technology now.

Physical AI Is Coming for Logistics

ChatGPT demonstrated what happens when AI understands language.

Image generators demonstrated what happens when AI understands visual patterns.

Autonomous vehicles demonstrate something else.

AI understands the environment and acts inside it.

That is Physical AI.

The system receives information from the real world.

It reasons.

It makes a decision.

Then a physical machine moves.

That creates a fundamentally different level of responsibility.

If ChatGPT gives a bad answer, someone may receive incorrect information.

If an autonomous truck makes a bad decision at highway speed, the consequences can be physical.

That is why verification, redundancy, simulation, regulation, and safety engineering become critical.

The Bigger Story

Gatik's $200 million funding round is important.

But the 85,000 reported driverless orders and $600 million in contracted revenue may be even more important.

They indicate that autonomous freight is moving away from a decade dominated by promises, prototypes, and demonstrations toward actual commercial deployment.

Gatik began with short, fixed routes.

Now its vehicles can operate across dynamic commercial networks involving highways, city roads, and dozens of pickup and delivery points.

Today there are dozens of trucks.

The company wants more than 100 by the end of 2026.

Eventually, it says it intends to operate thousands.

That is when this becomes much more than an autonomous vehicle experiment.

It becomes a new logistics infrastructure.