Daimler Truck and Torc Robotics Select Innoviz Technologies as LiDAR Partner for Series Production of Level 4 Autonomous Trucks

by Torc | Dec 2, 2025 | Featured, News

Daimler Truck North America and Torc Robotics Select Innoviz Technologies as LiDAR Partner for Series Production of Level 4 Autonomous Trucks

This press release has been released jointly with Torc, Daimler Truck North America, and Innoviz.

InnovizTwo High-Performance Short-Range LiDAR to enable SAE Level 4 autonomous capabilities for commercial trucks deployment

TEL AVIV, Israel: PORTLAND, Ore. and BLACKSBURG, Va. – December 2, 2025 – Innoviz Technologies Ltd. (NASDAQ: INVZ) (the “Company” or “Innoviz”), a leading Tier-1 direct supplier of high-performance, automotive-grade LiDAR sensor platforms and complementary software stack, announced today that Daimler Truck, one of the world’s leading commercial vehicle manufacturers and Torc Robotics, a subsidiary of Daimler Truck, have selected Innoviz as its Short-Range LiDAR supplier for series production SAE Level 4 autonomous Class 8 semi-trucks.

This announcement follows Innoviz’s previous disclosure that a major commercial vehicle OEM had selected the Company for future series production of Level 4 autonomous trucks, now revealed to be Daimler Truck. Innoviz will supply its InnovizTwo Short-Range LiDAR sensors to support Daimler Truck and Torc Robotics’ autonomous commercial vehicle program. As part of a joint development effort, the companies will collaborate to advance the sensors for commercial trucking applications.

Daimler Truck and Torc Robotics plan to integrate Innoviz’s LiDAR technology into the autonomous Freightliner Cascadia in combination with Torc’s virtual driver as one of several key components enabling Level 4 autonomous trucking.The partnership positions Innoviz’s technology as a critical component in Daimler Truck’s strategy to bring autonomous trucks to market, with deployment planned across highway and regional routes in North America to help fleet operators improve operational efficiency, and enhance road safety.

“This partnership with Daimler Truck and Torc represents a significant validation of our technology and our position in the autonomous trucking market,” said Omer Keilaf, CEO and Founder of Innoviz. “The trucking industry demands LiDAR sensors that can perform reliably in the most challenging conditions while delivering the precision and range needed for safe autonomous operation. Our InnovizTwo sensors have demonstrated compliance with these stringent requirements, and we’re excited to support Daimler Truck and Torc in bringing this transformative technology to market.”

“Selecting the right LiDAR partner is fundamental to our autonomous trucking strategy,” said Rakesh Aneja, Head of Corporate Development at Daimler Truck North America. “Innoviz’s proven track record in automotive-grade LiDAR sensors makes them an ideal partner as we advance toward series production. This collaboration brings us closer to delivering autonomous trucks that will reshape the logistics industry.”

Mike Avitabile, Head of Engineering at Torc, adds: “Integrating Innoviz’s technology into our self-driving vehicle software solution enhances our system’s ability to detect, classify, and track objects in real time across diverse road and weather conditions. Innoviz’s sensors deliver the consistency and durability required for commercial operation, while supporting the redundancy needed for safe Level 4 autonomy.”
Daimler Truck and Torc rely on a combination of three complementary sensor technologies – state-of-the-art LiDAR, radar, and camera systems – to precisely detect the vehicle’s surroundings under all conditions. This multi-layered approach enhances road safety both on highways and in challenging maneuvers such as turning at intersections or navigating ramps. LiDAR technology uses laser pulses to generate high-resolution 3D maps of the environment. For maximum safety, autonomous trucks require both long-range LiDAR systems to identify objects far ahead and short-range sensors to capture detailed close-proximity data in complex driving situations. While Daimler Truck and Torc have already selected their supplier for long-range LiDAR, Innoviz has now been chosen as the partner for short-range LiDAR.

About Daimler Truck

Daimler Truck Holding AG (“Daimler Truck”) is one of the world’s largest commercial vehicle manufacturers, with over 40 main locations and more than 100,000 employees globally. Daimler Truck North America LLC (DTNA), a subsidiary of Daimler Truck and headquartered in Portland, Oregon, is the largest manufacturer of heavy-duty trucks in North America and a leading provider of innovative products, services, and technologies for the commercial transportation industry. DTNA designs, engineers, manufactures, and markets medium- and heavy-duty trucks, school buses, vehicle chassis, and related technologies and components under the Freightliner, Western Star, Thomas Built Buses, Freightliner Custom Chassis Corp, and Detroit brands. DTNA is dedicated to delivering exceptional value and support to its customers—helping them keep the world moving. For more information, visit northamerica.daimlertruck.com or daimlertruck.com.

About Torc

Torc, headquartered in Blacksburg, Virginia, is an independent subsidiary of Daimler Truck AG, a global leader and pioneer in trucking. Founded in 2005 at the birth of the self-driving vehicle revolution, Torc has over 20 years of experience in pioneering safety-critical, self-driving applications. Torc offers a complete self-driving vehicle software and integration solution and is currently focusing on commercializing autonomous trucks for long-haul applications in the U.S. Torc has offices in Montreal, Ann Arbor, Blacksburg, and the Dallas–Fort Worth (DFW) area. Headquartered in Blacksburg, Virginia, Torc also operates an engineering office in Montreal, a fleet operations facility in DFW to support productization and commercialization efforts, and a facility in Ann Arbor to leverage the region’s strong autonomous and automotive talent base. Torc’s purpose is driving the future of freight with autonomous technology. As the world’s leading autonomous trucking solution, we empower exceptional employees, deliver a focused, hub-to-hub autonomous truck product, and provide our customers with the safest, most reliable, and cost-efficient solution to the market.

About Innoviz

Innoviz is a global leader in LiDAR technology, serving as a Tier-1 supplier to the world’s leading automotive manufacturers and working towards a future with safe autonomous vehicles on the world’s roads. Innoviz’s LiDAR and perception software “see” better than a human driver and reduce the possibility of error, meeting the automotive industry’s strictest expectations for performance and safety. Operating across the U.S., Europe, and Asia, Innoviz has been selected by internationally recognized premium car brands for use in consumer vehicles as well as by other commercial and industrial leaders for a wide range of use cases. For more information, visit innoviz.tech.

Join the discussion: Facebook, LinkedIn, YouTube, X

Media Contact
Media@innoviz-tech.com

Investor Contact
Investors@innoviz-tech.com

Forward Looking Statements

This announcement contains certain forward-looking statements within the meaning of the federal securities laws, including statements regarding the services offered by Innoviz, the anticipated technological capability of Innoviz’s products, the markets in which Innoviz operates, Innoviz’s projected future operational and financial results. These forward-looking statements generally are identified by the words “believe,” “project,” “expect,” “anticipate,” “estimate,” “intend,” “strategy,” “future,” “opportunity,” “plan,” “may,” “should,” “will,” “would,” “will be,” “will continue,” “will likely result,” and similar expressions. Forward-looking statements are predictions, projections and other statements about future events that are based on current expectations and assumptions and, as a result, are subject to risks and uncertainties.

Many factors could cause actual future events, and in the case of our forward-looking revenues, actual orders or actual payments, to differ materially from the forward-looking statements in this announcement including but not limited to, the ability to implement business plans, forecasts, and other expectations, the ability to convert design wins into definitive orders and the magnitude of such orders, the ability to identify and realize additional opportunities, potential changes and developments in the highly competitive LiDAR technology and related industries, and our expectations regarding the impact of the evolving conflict in Israel to our ongoing operations. The foregoing list is not exhaustive. You should carefully consider such risk and the other risks and uncertainties described in Innoviz’s annual report on Form 20-F for the year ended December 31, 2024, filed with the U.S. Securities and Exchange Commission (“SEC”) on March 12, 2025 and in other documents filed by Innoviz from time to time with the SEC. These filings identify and address other important risks and uncertainties that could cause actual events and results to differ materially from those contained in the forward-looking statements. There can be no assurances that the Company will enter into definitive agreements, orders or receive payments with respect to the program selection referenced in this announcement. Forward-looking statements speak only as of the date they are made. Readers are cautioned not to put undue reliance on forward-looking statements, and Innoviz assumes no obligation and does not intend to update or revise these forward-looking statements, whether as a result of new information, future events, or otherwise. Innoviz gives no assurance that it will achieve its expectations.

Torc Provides Fast, Secure Self-Service for Virtual Development Using Amazon DCV

by Torc | Nov 5, 2025 | Technology, News

This case study was originally posted at the AWS Solutions site.

Overview

Torc Robotics (Torc) wanted to facilitate remote development for its distributed workforce. The company develops autonomous vehicle software and technology that’s aimed at commercializing autonomous semitrucks by 2027. To support these efforts, Torc needed a secure, robust virtual desktop infrastructure (VDI) solution for engineers to run large GPU- and CPU-based workloads.

Torc, which was already using Amazon Web Services (AWS) for many of its workloads, built a VDI solution using Amazon DCV, which delivers high-performance remote desktop and application streaming. Now, Torc engineers have secure, highly available access to the compute resources that they need in minutes, and the company can continue working toward its goal of making highways safer using autonomous driving technology.

Opportunity | Using Amazon DCV to create the VDI Ranch for Torc

Torc—founded in 2005 and an independent subsidiary of Daimler Truck since 2019—is focused on delivering an autonomous trucking software product for hub-to-hub transportation, with the vision to provide fleet customers with the safest, most reliable, and cost-effective solution on the market. “Safety is a top priority at Torc,” says Jason Fox, senior engineering manager at Torc. “The trucking industry is facing driver shortages and inefficiencies, and there are many crashes on public roads that involve trucks. There is an opportunity to improve road safety and efficiency in freight transportation and Torc’s role in this is developing autonomously driving semitrucks.” In 2024, Torc completed validation of its first driver-out product release on production-intent hardware and software. The company is now testing on public roads from its autonomous hub in the Dallas–Fort Worth area.

Torc’s engineers and developers work from many locations, and the company sought to support remote development in a governed, standardized environment where it could secure its intellectual property. Torc also wanted to provide flexible access to GPU resources for the machine learning research and training that supports its autonomous driving software. At the same time, Torc did not want to create a centralized environment that would have high maintenance overhead or single points of failure. “We’re cloud engineers, so we think that things should be horizontally scaled, resilient, automated, and repeatable hundreds of times; not centrally managed or where a single developer’s issues will affect other people,” says Fox.

Torc tested various VDI solutions. As a customer of AWS since 2020, it looked to see what AWS had to offer. “We lean on AWS heavily for managed services whenever we can so that we can think more about writing code and making the trucks work,” says Fox. “The services that AWS offers made sense for this project as well.” Torc worked with the AWS team to test Amazon DCV. The solution worked well for the company, and Torc ultimately used it as the main component of its in-house VDI solution, the VDI Ranch.

Solution | Spinning up GPUs in under 5 minutes using Amazon DCV

The main principle behind the VDI Ranch is the ability to spin up and down instances as needed. “We strongly feel that in cloud computing environments, servers should be cattle, not pets,” says Fox. “We should have easily reproducible servers in the cloud, and when there’s a problem with a server, you delete it and spin up another. You don’t feed and care for it like a pet.” In fact, one of the options in the VDI Ranch is a “Replace Instance” button. If a server has an issue, the developer can simply replace the instance with a new one, keeping their data and settings intact.

The VDI Ranch provides a self-service, end-user compute environment for nearly 300 developers and engineers who can get access to the compute resources they need in under 5 minutes—rather than submitting a ticket and waiting several days to have resources allocated. This greatly accelerates developer productivity.

With the VDI Ranch, Torc can provide developers with flexible access to GPU and other high-powered computing resources using Amazon Elastic Compute Cloud (Amazon EC2), which provides secure and resizable compute capacity for virtually any workload. “Using AWS and Amazon DCV is a much easier way for us to provide GPU horsepower to developers when they need it,” says Fox. “We cannot provide laptops or even desktops with the kind of GPU power that we get from Amazon EC2 instances, and it’s flexible, so we can tear the instance down when we don’t need it anymore.” The VDI Ranch now powers every major area of Torc’s software development.

Torc implemented automated governance and security controls within the VDI Ranch, including integrating the VDI Ranch with Torc’s third-party identity and access management solution. Torc also implemented observability dashboards in Datadog to track networking and compute instance performance. These dashboards are used by the cloud engineers supporting the VDI Ranch, which has helped Torc more easily troubleshoot technical issues among its remote workforce, improving performance and latency.

The VDI Ranch also makes it possible for the Torc cloud engineering team to standardize the hardware that Torc employees use—which improves security and troubleshooting—while still giving engineers a development environment that uses their preferred operating system. For contractors, Torc uses Amazon WorkSpaces, which provides fully managed virtual desktops. “Using Amazon WorkSpaces, we get the benefits of managed VDI, including segmentation between employee and contractor workloads, and don’t have to manage Windows images,” says Fox.

Outcome | Improving remote development using AWS

As the company works toward releasing its autonomous trucks, Torc will continue improving the user experience of the VDI Ranch for its developers. It has recently deployed a system that intelligently shuts down instances that aren’t being used and has built a VDI-specific compute optimizer into FinOps dashboards to help users rightsize their compute resources. These measures will lead to better optimization and lower costs.

“This project would not have been possible without the AWS team engaging with us for the last 2 years,” says Fox. “I can’t think of a better relationship with a vendor who understands our challenges and helps us find solutions.”

“Using AWS and Amazon DCV is a much easier way for us to provide GPU horsepower to developers when they need it.”

Jason Fox

Senior Engineering Manager, Torc

Rebeca Delgado Joins Torc As Vice President, Engineering – Autonomy Applications

by Torc | Sep 16, 2025 | People Ops, Featured

Rebeca Delgado, VP Engineering – Autonomy Applications

BLACKSBURG, Va – September 16, 2025 – Torc Robotics, an independent subsidiary of Daimler Truck AG and a pioneer in commercializing self-driving vehicle technology, today announced the addition of Rebeca Delgado as VP, Engineering – Autonomy Applications, which includes responsibility for the Feature and Model Development groups.

An accomplished technology leader with more than two decades of experience in semiconductors, edge computing, and automotive systems, Rebeca most recently served as Chief Technology Officer and Senior Principal AI Engineer at Intel Automotive, where she led a diverse, multidisciplinary team in the Automotive CTO Office. In this role, she defined the company’s “whole vehicle” compute strategy, drove AI and high-performance compute innovation, and played a pivotal role in acquisitions and industry standards. Her team was responsible for pathfinding and incubation of next-generation vehicle systems and solutions, blending deep technical expertise with strategic vision and cross-functional leadership.

Rebeca’s career is marked by a steadfast passion for innovation at the edge of compute in vehicles and intelligent systems. She has been instrumental in shaping edge high-performance compute products, crafting roadmaps for autonomous driving, ADAS, and MaaS platforms, and tailoring solutions for global OEMs and Tier 1s. With expertise spanning software, compute architecture, and automotive technology, she has consistently driven industry standards and advanced the future of software-defined vehicles. A recognized thought leader, she frequently speaks at global conferences on mobility, intelligent vehicle platforms, and the evolution of AI-enabled automotive compute. Rebeca was also honored by Automotive News as one of the 100 Leading Women in the North American Auto Industry, highlighting her impact and influence across the sector.

She will be based in the Torc Engineering Technology Hub in Ann Arbor, Michigan.

About Torc Robotics

Torc, headquartered in Blacksburg, Virginia, is a global leader and pioneer in trucking. Founded in 2005 at the birth of the self-driving technology, Torc has 20 years of experience in pioneering safety-critical, self-driving applications. Torc is working toward a complete self-driving vehicle software and integration solution and is currently focusing on commercializing autonomous trucks for long-haul applications in the U.S. In addition to its Blacksburg headquarters and engineering offices in Austin, Texas, and Montreal, Canada, Torc has a fleet operations facility in the Dallas-Fort Worth area in Texas, to support the company’s productization and commercialization efforts, as well as a presence in Ann Arbor, MI, to leverage the autonomous and automotive talent base in that region. Torc’s purpose is driving the future of freight with autonomous technology. As the world’s leading autonomous trucking solution, we empower exceptional employees, deliver a focused, hub-to-hub autonomous truck product, and provide our customers with the safest, most reliable, and cost-efficient solution to the market.

Contacts

Laura Lawton, 408-505-5820
press@torc.ai

New Phase of Torc–Edge Case Collaboration Targets Production-Ready Safety Case

by Torc | Aug 19, 2025 | Featured, Safety

Sunny drone shot of the Torc Ann Arbor office location

Independent safety assessments by Edge Case mark a pivotal step in Torc’s journey toward commercializing Level 4 autonomous trucking

Blacksburg, VA — August 19, 2025 – Torc, a pioneer in commercializing self-driving class 8 trucks, today announced a new strategic collaboration with Edge Case (EC), frontier technologies and safety-critical systems experts. This next phase of collaboration will support Torc’s mission to fully commercialize Level 4 autonomous trucks for long-haul applications in the U.S., ensuring its driverless safety case aligns with the applicable AVSC Best Practices and guidance from the Open Autonomy Safety Case (OASC). This alignment will result in a more streamlined, well-structured, safety case that improves clarity, accelerates development, and enhances cross-functional review.

This initiative will focus on a series of independent assessments of Torc’s Driverless Safety Case Framework and Evidence Sufficiency Criteria, reinforcing Torc’s commitment to safety and independent validation in preparation for production and commercialization. Looking ahead, Edge Case will conduct an assessment of completed safety case evidence. These assessments will be conducted independently by Edge Case and will include detailed reports and collaborative review sessions with Torc’s safety, engineering and operations teams.

“Edge Case brings world-class expertise in building rigorous and comprehensive safety programs,” said Jerry Lopez, Senior Director of Safety Assurance at Torc. “Their leadership and experience across multiple autonomy segments make them an ideal partner as we move toward production readiness.”

This announcement comes on the heels of Torc’s recent appointment of Steve Kenner as Chief Safety Officer and the company’s ongoing prioritization of safety through integrated, cross-functional collaboration.

“This partnership with Torc represents a pivotal step forward in advancing autonomous trucking safety,” said Nathan Parker, Chief Executive Officer of Edge Case. “By leveraging our deep experience across autonomy domains, we’re helping ensure that Torc’s safety case is not only rigorous and transparent, but also production-ready for real-world deployment.”

With this collaboration, Torc is advancing toward its goal of launching fully driverless, commercial autonomous trucks for long-haul applications in the U.S. by 2027.

About Torc

Torc, headquartered in Blacksburg, Virginia, is a global leader and pioneer in trucking. Founded in 2005 at the birth of the self-driving technology, Torc has 20 years of experience in pioneering safety-critical, self-driving applications. Torc is working toward a complete self-driving vehicle software and integration solution and is currently focusing on commercializing autonomous trucks for long-haul applications in the U.S. In addition to its Blacksburg headquarters and engineering offices in Austin, Texas, and Montreal, Canada, Torc has a fleet operations facility in the Dallas-Fort Worth area in Texas, to support the company’s productization and commercialization efforts, as well as a presence in Ann Arbor, MI, to leverage the autonomous and automotive talent base in that region. Torc’s purpose is driving the future of freight with autonomous technology. As the world’s leading autonomous trucking solution, we empower exceptional employees, deliver a focused, hub-to-hub autonomous truck product, and provide our customers with the safest, most reliable, and cost-efficient solution to the market.

About Edge Case

Edge Case is the trusted technical partner for companies working with frontier technologies and sophisticated systems. Headquartered in Pittsburgh, Pennsylvania, Edge Case supports industry leaders across automotive, aerospace, defense, energy, and AI as they design, build, and deploy complex, critical systems. With deep expertise in autonomy, functional safety, and systems engineering, Edge Case helps teams navigate evolving regulatory landscapes and operational risk. Utilizing DevSafeOps, Edge Case enables organizations to digitize safety workflows, generate defensible safety cases, and accelerate readiness for launch. Whether supporting autonomous vehicles, robotic systems, or next-generation AI applications, Edge Case is on a mission to ensure a safer tomorrow.

Contacts

Laura Lawton, 408-505-5820
press@torc.ai

AV 3.0: Torc’s AI Blueprint

by Torc | Aug 6, 2025 | News, Education

The autonomous ready Freightliner Cascadia, with Torc virtual driver software embedded within the chassis.

Torc has set our sights on one goal: to create a scalable product capable of delivering the immense value of autonomy to our freight partners. The system, hardware, and software must all be built for safety, adaptability, and cost-efficiency; and work seamlessly on Daimler Truck’s Freightliner Cascadia. And Torc has cracked the code: we’ve designed an unparalleled system that unlocks safe and reliable autonomous driving for long-haul trucking, which can be quickly scaled to meet the needs of our fleet customers. We call it AV 3.0.

AV 3.0

To hit the high goals we set, we recognized that today’s most advanced AI approaches being used to develop autonomy today would not meet the level of capability, performance, and most importantly, safety that the trucking industry requires. We knew we needed to achieve several firsts for autonomous vehicles in parallel: a new autonomous driving architecture as well as a new data and development infrastructure. Torc tapped our world-class AI talent and our long history in robotics and self-driving vehicles to create AV 3.0: our virtual driver software, the advanced data loop, and generative AI simulation infrastructure to train and test it, setting ourselves apart from others in the industry. Let’s break down the parts and explain the advanced components of each.

The Software Product: Virtual Driver

Virtual driver, our autonomous software product driving the truck, is comprised of three main functional modules: perception, prediction, and planning. It is an end-to-end reinforcement-learned (RL) modular software stack with redundant rule-based (heuristic) guardrails. This allows us to uniquely deliver the optimal mix of autonomous driving performance, system reliability, and driving safety. While the machine learning modules in the stack provide the highest driving performance output, Torc layers in rule-based models and arbitration as redundant guardrails, ensuring the ML model doesn’t violate the basic rules of driving.

The addition of the heuristic guardrails is one of the benefits of Torc’s 20-year heritage in developing autonomous vehicle software products. Building rule-based models is an intensive, manual, and time-consuming process, and Torc has been working with heuristic rules as part of our code since the beginning. Also, the addition of rule-based models is inherently verifiable and interpretable since the rules are hand-coded – another layer of verifiability to support a strong safety case.

The modular architecture of the stack provides fine-grained introspection: the ability to see and validate with transparency and explainability if all the learned and heuristic modules are reacting properly to the input. In comparison, an AV 2.0 end-to-end system is a “black box” with just sensor data in and trajectory out. You can’t understand how the stack is reasoning, only that it is reasoning. There’s no way to check on the intermediary results to check if the module is behaving properly and therefore, no way to effectively debug issues. Every time an issue is found, an AV 2.0 end-to-end system must be retrained as a whole with new data specific to that issue to achieve a correct output, but there is no way of knowing if it truly fixed the issue. This “whack-a-mole” approach adds significant time and uncertainty to the software creation and safety validation process.

With our AV 3.0 stack, we can understand how the system is reasoning within each module, in perception, prediction and planning, and where the problems are, retrain the stack with the right set of generated data to specifically target the issue, and supervise the module-level output to validate the fix.

Perceived object bounding boxes and lane centers from the perception module.

The Simulation Environment: Learning to Create the World

Supporting the training of our virtual driver is our world generation and data-driven simulation environment, comprised of several important and complex pieces.

It Always Starts with the Data. We capture real-world data with the actual production-intent sensor suites on our trucks and have collected a breadth of highway and surface street data across many regions and driving conditions.

In fact, Torc has by far the largest set of real-world driving data in autonomous trucking, spanning many diverse scenarios and corner cases.

We then recreate that real-world data within our data-driven simulation environment, giving us the ability to replay those scenes ad infinitum to test the virtual driver. But that real-world data, used repeatedly on replay, is also the starting point for the generative AI environment we’ve developed, to create infinite sequences of challenging driving scenes and scenarios necessary to train and test the virtual driver to meet the needs of our ODD and safety goals.

As no such capability to generate the realistic data needed was available in the market, the Torc AI team leveraged years of research and development to deliver these capabilities, providing us with a unique time, cost, and data fidelity advantage.

In fact, Torc has by far the largest set of real-world driving data in autonomous trucking, spanning many diverse scenarios and corner cases.

These can be summarized as the following three capabilities:

Neural Rendering

A generative AI approach for realistic scene creation. Based on real world scenes, we can recreate freight-specific, customer-desired sections of road with various “actors” — cars, trucks, animals, pedestrians — and then generate unique additions and/or changes to the scene, like a car braking hard in front, a wrong way driver, a light snow shower that becomes a blizzard, in a photo-and-lidar-realistic output that can be injected into our perception stack.

World Simulation

This approach generates the other reactive objects that react to our truck (the ego vehicle) and then manages the behaviors however we want or need. We can make reactions more or less aggressive on all of the actors, creating infinite scenes and behaviors. For example, one car captured behaving normally on one section of road can then be made to behave aggressively by degrees. It can go from benign to aggressively merging into the truck’s lane in various places by millimeters and timings by milliseconds.

Importantly, all learned behaviors are grounded in real-world physics-based models, so while we could test with Hollywood-style simulations of flying cars or people, we stay grounded in actualities. We instead simulate vehicles skidding into our lane where in the real-world data, they had been driving carefully in their own lanes.

scenario generation

Scenario generation takes world simulation to the next level, generating never-before-seen road geometries, creating all road conditions, all edge cases, and repeating the simulations through millions of miles, allowing for an even more diverse set of road scenarios. This allows us to quickly adapt the virtual driver software to any new route with minimal additional data collection and time, dramatically accelerating our ability to scale our software to drive new highways, new autonomous hub layouts, in different states, and so on.

Neural Rendering cropping

Real world footage

“For fully data driven simulation, you need to be able to simulate the behaviors and do so very realistically. You also need to be able to control the reactions, in a data-driven way,” says Felix Heide, Head of AI at Torc, and one of the architects of Scenario Dreamer™, a recent diffusion method for scenario generation. “This combination of world simulation systems accomplishes what is necessary.”

The Proof: Training and Testing

Using these layers of world simulation, neural scene rendering together with generated sequences with fully reactive agents, we repeatedly fine-tune and deeply test the virtual driver, as mentioned previously, an instructor teaching a “behind-the-wheel” student, by using reinforcement learning, to explore decision making and provide feedback at the same time.

Virtual driver is taught not only by providing the ideal road trajectory, but also by providing additional scene information so that the system learns to explicitly predict and, on top of that, to explicitly use the information for its final trajectory decision. This, in return, allows a more accurate training process, which, as a byproduct, helps us understand in more detail how the entire system behaves. That detail can then be fed back into the system.

We apply all these techniques to create the data needed to deeply test the virtual driver behavior for the most unusual or unlikely corner cases, to give us high confidence of safe behavior on the road, i.e., that the training of our AI system and the virtual driver has been successful. After that confidence threshold is reached, we move the updated model out of the simulated environment and onto the trucks, validating on both closed course and public roads for every software release.

It’s this high-velocity AV 3.0 framework of data –> training –-> testing –> release that sets an industry benchmark in software development and deployment.

The End Result: Safe and Scalable

“End-to-end systems are critical for scalability. It allows us to be much more adaptable to new needs or to fix issues quickly, and because we have a modular transparent stack, we can see the issues independently,” says C.J. King, Torc Chief Technology Officer. “We can use our generative AI stack to generate the data to train the outcomes rather than train a camera-based object and distance detector, or lidar and radar object detection separately from the prediction and planning. It’s another layer of getting the technology ready for the road.”

This high velocity is critical not only to keep to our release schedule, but to ensure we can quickly enable our customers to scale to new routes and hubs. We can identify, triage, and deploy issues to engineering before we ever encounter them on the road. Additionally, issue resolution and fixes are done within days, not weeks. This turnaround means there can be multiple fixes and releases daily.

Why We Call It AV 3.0

Why do we call it AV 3.0? Torc’s innovation is the next step beyond the technology in AV 1.0 and AV 2.0, as we knew those approaches alone weren’t enough to succeed in creating and deploying autonomous trucks in a timely manner. Torc’s AV 3.0 drives the state of the art in the evolution of scalable, practical, and applied autonomous vehicle software. Until we reach our product launch in 2027, we will continue to hone this technology, gather more real-world data (something we will always need), develop commercialization plans and hub designs, close our robust safety case, and solve a multitude of other necessary autonomous trucking puzzles. But AV 3.0 provides a clear path to how it can be done. So, what’s next? We’ve defined AV 3.0, the groundbreaking technology used to create, train, and test the virtual driver. What our virtual driver runs on, and how, is another industry-defining element setting Torc apart from the competition. We’re deploying this technology right now, not on prototypes or demonstration trucks, but on the trucks we will go to market with… on production-intent embedded hardware.

Next Up: “Applied AV 3.0: The Key of Embedded Hardware ”

Transparency

AV 3.0 understands the state of the entire model and why it acted the way it did, both within individual modules and how it affected the entire end-to-end system. There’s visibility and understanding of what’s always going on inside the model at all times.

Explainability

AV 3.0 permits traceability and verifiability, providing us with the ability to see why and where our virtual driver did something. We can understand both what is detected and what reasoning the system made to determine the object and its meaning. This is crucial for the explainability of our end-to-end system and for our safety case.

Improved Accuracy

Introspection means the AI model is aware of its internal state, how it makes decisions, and its behaviors. Models can understand why they are doing something and can therefore be more accurate.

Let’s explain EXPLAINABILITY and IMPROVED ACCURACY together: AV 3.0 provides object detection outputs and information on the 3D world view our vehicle is in – Is that a car? At what velocity are other highway users travelling at? Is that an emergency vehicle? Where is the lane center? A traffic light? This level of detail allows us to understand what the virtual driver thinks in a certain situation, and we also know how it acts with that information.

Improved Safety

With full transparency of the entire model and individual modules, it is easier to monitor model behavior and verify the software from the inside and explain perception and decision making in full.