Proven, Not Promised: Earning Our Place on the Road

Proven, Not Promised: Earning Our Place on the Road

Torc truck on Texas highway

At Torc, safety isn’t just a priority; it’s the foundation that supports every aspect of how we develop, deploy, and operate our autonomous trucking technology. As we work to transform commercial freight transportation, we’ve built an approach that establishes trust with regulators, our customers, and the public, that Torc’s autonomous solutions are safe to deploy and in the communities we operate in.

Torc’s Chief Safety Officer, Steve Kenner, and his team have developed a safety framework organized around three pillars that guide our safety philosophy. Individually, they keep different parts of our whole team accountable to specific segments, such as the engineering of our trucks, how they’re operated, and how safety is fostered within Torc. Together, they ensure we’re not just building a safe autonomous truck but operating it responsibly.

Build it Safe

The foundation of our safety approach starts with the truck itself. We’ve engineered multiple layers of redundancy into our hardware as well as our autonomous driving systems, embedded on the industry-leading Freightliner Cascadia platform, to ensure safe operation even when individual components fail.

  • Redundancy at Every Level

Our redundancy approach encompasses all critical systems. Our safety philosophy spans all aspects of our technology, from our system architecture to our hardware and software design, allowing us to be resilient to failures. This is accomplished through redundancies in our safety-critical systems, such as braking and steering.

For instance, if an individual sensor fails, we can achieve an MRC (minimal risk condition) by pulling off the road at a safe location. The redundancies we’ve built in allow us to continue operating safely.

  • Intelligent Fault Management

Torc has developed a sophisticated fault-management system that detects faults, including safety-relevant faults, and matches them to appropriate responses. Our priority hierarchy is clear: first and foremost, our trucks must maintain safe operations and refrain from creating a hazard for other road users. If we can’t reach our destination, we focus on degraded operation that gets us to a safe location, rather than just pulling to the roadside. Think truck stop or exit ramp, not highway shoulder. However, in safety-critical situations, our autonomous truck will pull over to the side of the road as soon as it is safe to do so.

  • Addressing Edge Cases

Edge cases (problems or situations that occur at the extreme limits of a system’s operating parameters) are inevitable in the real world. Whether it’s a billboard featuring a giant stop sign that could be misidentified by sensors, or a pickup truck loaded with Christmas trees, our autonomous driving system needs to handle a multitude of scenarios not previously encountered. Torc is tackling this challenge through a collaboration with Stanford University, where we’ve partnered to make the future of freight safer for all. By sharing our data and perception information with our partners, we’re able to compare datasets and identify differences between them. For instance, let’s say our simulation training data includes lots of pickups loaded with wooden planks, but we don’t have any of those same “real world” vehicles in our on-road perception data. These kinds of insights allow us to identify areas where we should develop a more complete perception dataset.

We also leverage publicly available crash databases from the National Highway Traffic Safety Administration and state agencies to analyze crashes on our planned routes (as well as roads outside of our routes) but within our ODD or operational design domain, which the set of specific conditions and locations under which an automated system can/is allowed to operate. This allows us to replay and recreate these scenes during software testing, using AI generated scenarios in our simulation environment. We can test and train the software’s responses to both real-world crashes and even more challenging simulated situations.

    • The Role of Simulation and Testing

    Let’s dive deeper into Torc’s simulation environment and our testing and training. Torc firmly believes that you cannot real-world-road-test our way to safety. Even millions of miles on the road wouldn’t be enough to encounter all potential edge cases. Therefore, we use a balanced combination of (1) real-world data and (2) generated AI simulated scenarios to test our software, and then (3) retraining the system as a whole. We use safety metrics, such as hard breaking, hard swerves, and total time to collision that allow us to measure performance in both environments and establish parity between simulation and on-road results.

    Our comprehensive simulation libraries provide scalable, cost-effective testing capabilities, but we still need sufficient on-road miles to validate our software against real-world performance.

    Operate it Safe

    Technology is only part of the safety equation. With autonomous trucking, Torc is improving how freight operations work and realize efficiencies, prompting new approaches to maintenance, inspection, and human oversight:

    • Evolving Operational Standards. Our autonomous technology requires updated maintenance and inspection protocols compared to traditional trucking operations, such as sensor calibration and increased frequency. We’re developing the training programs and procedures that will equip fleet operators and technicians with the skills they need to work with autonomous vehicles.
    • Human Oversight Philosophy. We’re establishing clear guidelines as to when remote assistance should assist a vehicle when it encounters situations that may require human oversight. This balance ensures that human judgment enhances rather than undermines our autonomous capabilities.
    • Consistent Standards Everywhere. Safety standards must be maintained across different routes, terminals, and regulatory environments. We’re building systems that adapt to varying environments while maintaining our core safety principles.

    Trust It To Be Safe

    We recognize that the biggest challenge we face now isn’t necessarily technical; Torc has already overcome that. It’s earning public trust. We’re committed to establishing and maintaining safety from our fellow drivers, regulators, partners, and our community at large.

    • Transparency and Engagement

    As Torc CSO Steve Kenner puts it, “We’re all in charge of public safety.” We maintain transparency with regulators and elected officials who are responsible for public safety, engaging with stakeholders at city, county, state, and federal levels. We want people to understand what we’re doing, how we’re doing it, and when we’re doing it. This includes comprehensive first responder interaction plans and training programs, like these, ensuring that emergency personnel understand how to interact safely with our vehicles.

    • Independent Validation

    We understand that industry self-assessment isn’t enough. That’s why we’re working with third parties like Stanford’s Center for AI Safety and Edge Case Research; we’re actively seeking out neutral third-party studies and evaluations of our work to ensure that we’ve created a complete picture of our safety case.

    • Industry Collaboration

    Torc believes that rising tides lift all boats, so, while we compete on many aspects of our business, we should always collaborate on safety. We participate in industry consortiums and safety organizations, such as the Automated Vehicle Safety Consortium, sharing knowledge that advances autonomous vehicle safety across the entire sector, while respecting antitrust and intellectual property requirements.

    How It All Comes Together: Governance and Risk Management

    These three pillars of Torc’s safety approach don’t operate in isolation; they’re connected through a comprehensive governance structure centered on risk management. We’ve established a flexible safety governance framework where safety decisions can be promptly reviewed by a cross-functional team of experts. These committees report up to the Torc safety board, which CSO Steve Kenner leads.

    Safety risk management is pervasive throughout our approach. We systematically identify safety hazards and develop mitigation strategies across all our operations, whether we’re selecting hardware, planning operations, or making any other business decisions. Our safety risk management framework provides the structure for how we assess and evaluate these safety risks.

    Earning Our Place on the Road

    As autonomous trucking technology continues to mature and regulatory frameworks evolve, our three-pillar safety philosophy provides a comprehensive approach to the complex challenges of deploying self-driving vehicles at scale. We’re not just working to put autonomous trucks on the highway – we’re working to earn their place there through demonstrated safety, reliability, and responsible operation.

    The transformation of commercial transportation will succeed only if we can prove that autonomous vehicles operate safely alongside human drivers while delivering the efficiency and economic benefits they promise. Our systematic approach to safety, governance, and public engagement is designed to make that vision a reality.

    At Torc, safety isn’t just what we do; it’s who we are. Every decision we make, every system we build, and every mile we drive is guided by our commitment to operating safely and earning the trust of everyone who shares the road with us.

    Securing the Future: Meet Michael Maass, Director of Product Cybersecurity

    Securing the Future: Meet Michael Maass, Director of Product Cybersecurity

    At Torc, safeguarding safety-critical systems from evolving cybersecurity threats isn’t just a priority—it’s a core
    mission.

    At the helm of this mission is Michael Maass, the Director of Product Cybersecurity and Principal Product
    Cybersecurity Architect, whose career reflects a deep dedication to building secure technologies, strong teams,
    and forward-thinking strategy.

    A Career Rooted in Cybersecurity Excellence

    Michael brings over 20 years of cybersecurity experience to the table—17 of those spent specifically on
    securing companies, products, and safety-critical systems. His background bridges the technical and strategic:
    from writing low-level software in x86 assembly, C/C++, and Java, to leading cross-functional security teams
    and developing ways to build security into products.

    Throughout his career, Michael has helped companies adopt cutting-edge practices that meet both technical and
    compliance requirements. His passion lies in one of the most complex and important areas of modern
    technology: developing secure, safety-critical cyber-physical systems that can stand up to real-world threats
    and regulatory scrutiny.

    Building Secure Teams and Culture

    Michael’s leadership extends beyond technology. He’s spent years building and nurturing top-tier cybersecurity
    teams in the automotive space, ensuring they not only have deep technical skills but also understand the
    broader impact of their work. His ability to communicate with executives, regulators, and external stakeholders
    makes him a rare bridge between engineering and compliance—a necessity in today’s high-stakes
    cybersecurity environment.

    Leading Cybersecurity at Torc Robotics

    At Torc, Michael wears two hats: he acts as both Principal Product Cybersecurity Architect and the Director of
    Product Cybersecurity. In these roles, he leads efforts to embed cybersecurity into every stage of product
    development for autonomous vehicles—systems that must operate safely and securely in the unpredictable real
    world.

    His leadership ensures that cybersecurity is an integral part of a product’s full lifecycle, ranging from inception
    to product retirement.

    Penetration Testing: Red Teaming for the Right Reasons

    One of the key tools in Michael’s cybersecurity arsenal is penetration testing—a technique that simulates real-
    world attacks in order to find and fix vulnerabilities before bad actors can find and exploit them.

    “Penetration testing is essentially where someone with hacking skills applies those skills to a particular target,
    with the goal of efficiently finding vulnerabilities and identifying hardening opportunities,” Michael explains.

    Penetration testing is a process within the A Versatile Cybersecurity Development Lifecycle (AVCDL), an open
    source document set crafted by Charles Wilson, Cybersecurity Architect at Torc, Michael, and cybersecurity
    engineers at Torc and other companies, for use by the autonomous vehicle industry and any other creating
    safety-critical cyber physical systems. This structured lifecycle ensures that every aspect of the product, from
    hardware to software, is rigorously vetted for cybersecurity risks and those risks are addressed.

    Penetration testing is just one part of a broader cybersecurity strategy. While penetration testing is often
    spotlighted because it’s exciting and accessible, it’s important to note that a secure cybersecurity platform
    includes a comprehensive set of practices. Many of these, while equally critical, are less known unless you’re
    deeply immersed in the domain.

    Still, penetration testing stands out as an illustrative example of how Torc’s structured lifecycle ensures that
    every aspect of an autonomous vehicle product—from hardware to software—is rigorously vetted for
    cybersecurity risks.

    Inside the Penetration Testing Process

    As Michael says, no two penetration tests are the same, but most follow six general steps:

    • Pre-Engagement: Define the target, set objectives, and establish boundaries.
    • Reconnaissance: Gather data on the system using both passive and active techniques.
    • Threat Simulation: Emulate the tactics of real-world threat actors.
    • Exploitation: Attempt to breach the system, identifying weak points.
    • Analysis: Assess findings and potential business impact.
    • Reporting & Recommendations: Share results and collaborate on mitigation strategies.

    In a real-world example of a penetration test on a lidar system, Michael shared that safety and security go
    hand-in-hand. Both components must work together to ensure the strength of a system. During this particular
    lidar test, Michael pointed out that analog attacks, while harmful, aren’t necessarily as impactful as system-level
    exploits.

    Michael’s Vision for Secure Autonomy

    Michael’s ultimate goal is to create technology that’s secure, reliable, and compliant—all while enabling
    innovation. His work ensures that Torc’s systems are ready not just for today’s challenges, but that the larger
    autonomous vehicle industry is ready to face tomorrow.

    As Michael says, a rising tide lifts all boats. At Torc, that philosophy is core to how cybersecurity is
    approached—not just as a competitive advantage, but as a shared responsibility across the industry. That’s why
    Torc is pushing forward with tools like the AVCDL. By making this framework visible and accessible, Torc aims to
    help everyone build safer, more secure, and more compliant products.

    “I’m passionate about developing secure, safety-critical systems that balance innovation with acceptable
    liability and compliance,” he says. In an industry where trust is everything, that mission is more vital than ever.
    With decades of experience and a passion for securing the future of mobility, Michael Maass is helping to
    shape the next generation of cybersecurity in autonomous vehicles. Through strategic leadership, technical
    expertise, and a commitment to continual improvement, he’s ensuring that Torc stays ahead of the curve—
    keeping systems safe, secure, and ready for the road ahead.

    Growth and Relentless Focus: Setting Up 2025 to Realize Long-Term Success

    Growth and Relentless Focus: Setting Up 2025 to Realize Long-Term Success

    Torc has begun successful advanced validation of our autonomous trucks without a driver in a multi-lane closed-course environment.
    As our team continues to drive the future of freight and push strongly into the productization phase of our development cycle, Torc is building out new teams in both the Dallas/Fort Worth (DFW) area at our new autonomous hub location and Ann Arbor, Michigan, a hotbed of experienced software talent. The need for highly skilled workers in the fields of AI, machine learning, simulation, systems engineering, and data analysis has skyrocketed and Torc’s existing fully remote employee base in Michigan underscores its commitment to fostering a technology workforce capable of supporting revolutionary change in logistics and freight transportation. Additionally, Torc is shifting focus to Dallas/Fort Worth to realize our commercialization strategy timeline as we prepare to go live on our first launch lane between DFW and Laredo. “Torc continues to be in a period of growth and reinvestment, while consistently evolving the business at a rapid pace,” says Andrew Culhane, Chief Commercial Officer. “We’re being deliberate about where we spend our time, energy, and resources as we move from the R&D and advanced engineering phases of our development lifecycle, and into the productization phase.”

    2025 location strategy

    To create growth in the right areas and to attract the right talent, we need to strategically build our workforce in locations that best support our commercial business needs in the U.S. To that end, we are winding down and transferring operations and facilities from our original testing location in Albuquerque, New Mexico and technology hub in Stuttgart, Germany during the first half of 2025. Torc will be shifting many of its Albuquerque resources to Dallas and starting a hiring push there and in Ann Arbor, hiring over 100 positions in each location over the coming months.

    “The Albuquerque and Stuttgart locations were critical for our development phases, as we tested on New Mexico highways and developed specifications for our product in Stuttgart. As we shift focus to testing on our launch lane from DFW to Laredo and developing a world-class team, we need to locate our resources in the areas that align with our future plans and lay the foundation for scaling the business to our customers’ needs,” says Culhane.

    MICHIGAN

    Interested in joining the Torc team in Michigan now? We ask that all interested applicants first fill out the General Interest Application – Engineering Talent form to pre-screen for roles in specific, needed sectors. If we think you may be a good fit for Torc, we’ll be in contact. We’re offering roles such as:

    • Software Engineers (C++, Python, ROS, Linux, SIL)
    • Embedded Engineers (NVIDIA, Ethernet, CUDA, GPU Optimization)
    • Automotive Software Engineers (AUTOSAR BSW (Base Software), Embedded C, Vector tools.)
    • Machine Learning Engineers (PyTorch, TensorRT and OpenCV)
    • DevOps Engineers (AWS)
    • Engineering Leadership

    “We’ve set our sights on Ann Arbor because of its proximity to Detroit automotive and high-tech talent, and it makes good sense for our productization strategy to be in the Ann Arbor area as we grow and reinvest in our business,” says Meghan Juanarena, Director of Talent Acquisition at Torc. “We are also looking forward to tapping further into one of the nation’s top 10 robotics programs at the University of Michigan.”

    Attendees interested in automation, artificial intelligence, machine learning, and software development will find roles that align with their skills and ambitions.

    As part of the Daimler Truck family, we’re laser-focused on developing self-driving technology specifically for the Freightliner Cascadia, a challenge with unique obstacles and immense potential to boost our supply chains and keep our roads safer. This event is part of Torc’s larger commitment to building a team of talented, innovative individuals who are passionate about bringing the next generation of autonomous solutions to the road.

    Are Autonomous Vehicles the Future?

    Are Autonomous Vehicles the Future?

    The future of everyday technology is full of innovations taken straight out of The Jetsons. But from 3D-printed food to robot butlers, we’re not so far from that sci-fi idea of what tomorrow looks like. Today, we ask our smart assistants to play our favorite songs and program our robot vacuums to clean the house while we’re at work. Our future, and our present, is full of these kinds of forward-thinking inventions, including self-driving vehicles.

    While self-driving vehicles are still a ways out from becoming flying cars, autonomous transportation is rapidly becoming part of our current lives. In fact, many cars on the road currently have some degree of autonomy – and we’re not just talking about luxury autopilot features.

    Programs like lane-keep assist and cruise control are all considered Level One Autonomy, or assisted driving. These days, cars with Level One autonomous functions are very, very common; it’s likely that you’ve driven one yourself. And as for Level Two Autonomy, wherein vehicles control both steering and accelerating/decelerating, there are a few cars on the road with this function, albeit their numbers are rapidly rising as autonomous technology becomes more affordable.

    We might have autonomous vehicles now, but are autonomous vehicles the future? Will that Jetsons-esque future become our everyday?

    Let’s Explore: Autonomous Trucking

    At Torc Robotics, we’ve been in the autonomous technology world since 2005. As seasoned veterans in the self-driving world, we whole-heartedly believe that autonomous vehicles are the future. We believe that autonomous trucking will be one of the first forms of self-driving to make it to the market, with Torc trucks at the head of the pack.

    Autonomous trucks, like the ones we’re developing at Torc, are the most viable form of self-driving technology due to a host of reasons, but some of the most apparent are our trucks’ addition of lidar and radar in combination with a full camera suite. When added to a Class 8 truck, these features allow our vehicles to understand what’s in the surrounding environment in real-time.

    And while autonomous driving systems will certainly make the roads safer for all, autonomous trucking adds a layer of usefulness that personal self-driving cars aren’t geared towards. From strengthening supply chains to driving down the cost of goods, trucks like ours will create a better future in countless ways.

    So, are autonomous trucks the future? Yes! Once it’s safe for driver-out vehicles to put tires to the pavement, you can expect autonomous trucks to carry goods to-and-from shipping hubs across the country.

    Let’s Explore: Autonomous Cars

    In the past few years, autonomous cars have exploded onto the scene. Although there are some flashy headlines and sleek cars rolling out onto city streets, are autonomous cars the future?

    The answer is: yes, but it’s a little more complicated than the autonomous trucking equivalent.

    Autonomous cars are certainly probable and possible; we’ll likely see Level Four autonomy on our streets in the near future. Thanks to algorithmic precision and sensors far better than the human eye, these cars will make our roads safer. They’ll also be more efficient thanks to carefully managed gas consumption, eco-friendly decision-making, and more.

    However, the cost of hardware is currently a huge challenge for the kinds of cars that we park in our driveways. From high-performance computing to mass-market manufacturing, it’s difficult to create a self-driving car that is safe, affordable, and maintainable for the average person. That’s not to say that it’ll never happen – it’ll just take some time before technology catches up to affordability. For now, those looking for a self-driving vehicle can look towards the various Level Two autonomous cars currently available – or set their sights on the robotaxis that are currently coming to many cities across the United States.

    Other Considerations: Software, Weather, and More

    An autonomous vehicle, whether it’s a self-driving truck or a self-driving car, is an amalgamation of sensors, algorithms, processors, and more. At autonomous driving companies like Torc, we’re working to develop software that combines these elements in seamless, sophisticated ways. We’re laser-focused on developing trucks that operate autonomously in a few particular use cases: for example, in ideal weather conditions.

    You might’ve heard that autonomous driving systems are still tackling the inclement weather challenge. While autonomous vehicles use a variety of sensors, cold inclement weather is still something the industry has yet to solve. Therefore, self-driving vehicles will likely launch in the Sunbelt region of the United States, where the weather is mostly temperate and snow isn’t a concern. However, we don’t expect snow to be an issue for long; there are promising solutions on the horizon, including ground-penetrating radar and other forms of localization methods. With so many developers working in a fast-paced, innovative industry, it’s only a matter of time before today’s problems become tomorrow’s footnotes.

    So, Are Autonomous Vehicles the Future?

    Yes. Autonomous vehicles are the future, just as they’re part of our present. As the autonomous driving industry continues to race towards self-driving solutions, we’re excited to see where the future leads us.

    Why Level 5 Autonomy Is Science Fiction vs Science Fact

    Why Level 5 Autonomy Is Science Fiction vs Science Fact

    What’s the number one question Torc’rs get asked? “When will your trucks fully reach Level 5 autonomy?” Humans like to think in the binary. Either you’re driving the vehicle, or you’re not driving the vehicle. But it is not so black and white when it comes to autonomous vehicles.

    The SAE – the Society of Automotive Engineers International — has defined autonomous driving into six different levels. Only Level 5 is fully autonomous, to the point of theoretically being able to handle every driving situation that you could imagine, in every location in the world, without any human interaction or supervision needed.

    Within the parameters of the 21st century, the software needed to create fully deployable Level 5, not to mention commercially viable vehicles, are science fiction.

    Level 4, just a step under that, is reality. It is, in fact, the science versus science fiction.

    What’s the difference between Level 4 autonomy and Level 5 autonomy?

    The nuances between Level 5 and Level 4 are distinct and important. Understanding them is key to differentiating between the real and the fantasy.

    Level 5 autonomy is where most imaginations run to when they hear “self-driving.” It is fully and completely autonomous. There is no need for a steering wheel, and all people are passengers. But that simplicity of Level 5 would require an astronomical amount of technological development, maintenance, and testing.

    For a fully realized Level 5 instance, the autonomous driving system would have to navigate everywhere. Every drivable surface, from gravel roads to back-alleyways would need to be mapped and remapped. Situations that the vehicle would be able to handle would include everything from unmarked roads and new pop-up roadworks to a tornado dropping out of the clouds, as aliens land, and Godzilla starts attacking. The definition of Level 5 includes the autonomous vehicle making the right decisions in all those situations without complaint. In other words, an android would be driving the vehicle; there would be zero human intervention involved.

    It is science fiction. Does that mean it will not happen? Maybe. However, Level 5 autonomous trucks are not a viable, achievable, accessible product for the 4.19 million miles of U.S. roads. And those vehicles, and that freight market, is what Torc is focused on.

    Torc’s Focused Approach

    Level 4, however, stays just slightly over that line of science fiction into science fact. Torc’s autonomous driving system can navigate without human intervention within environmental, route, or traffic scenario limits. The limits put on the expected or allowed performance of the trucks are defined in the system’s Operational Design Domain or (ODD). Within these factors, the autonomous truck can navigate the road ahead in defined situations. Importantly, Level 4 can be an autonomous product that can be realistically foreseen, programmed, tested, and viable. That is Level 4, a high automation drive.

    We state the Whys to Level 4 Autonomous Trucking plan in our Safety Report:

    Torc is working toward building a level 4 truck because we believe that this level of driving automation will be the safest for heavy-duty trucks and will offer enormous advantages in many areas.

    Level 4 trucks will:

    • Make significant contributions toward safer roads and more sustainable transport solutions;
    • Enhance traffic safety thanks to a redundancy of systems and multiple sensors that can consistently achieve high performance – thereby helping to prevent the great majority of crashes that today result from human error;
    • Improve efficiency and productivity by, among other avenues, higher utilization of the vehicles – practically around the clock;
    • Avoid traffic jams by intelligent route management, making it possible to travel during light traffic times.

    Level 5 Autonomy is Always Everywhere

    Is Level 5 autonomy possible? The answer is complicated. For answers, we turned to Axel Gern, chief technology officer of Torc’s technology and software development center in Stuttgart, Germany. “Level 5 automation means, in my definition, always everywhere,” says Axel. “It needs to make safe decisions when you’re driving through the desert in Africa, or in the middle of Bangalore. But do we need to drive there autonomously? No, not right now.”

    “Level 5 autonomy puts a lot of work and focus on these things, to achieve a safe system that is able to drive everywhere, every time,” Gern continues. “Level 4, when you start to restrict the scope, understanding the limits and living within those limits, is perfectly fine. When you basically cover the situation to at least a safe condition, what is called minimal risk condition or MRC, that is what’s useful. Anything else is a huge challenge. Level 4 autonomy, in the way Torc defines it, is approachable and doable.”

    For the trucking and the freight logistics industry, Level 4 is the most appropriate and realistic. It is also, strategically, the safest level, and since the beginning of our company, safety and saving lives has been our work’s core mission.

    At Torc, our mission is to create a scalable, commercially viable autonomous trucking product that will address the practical needs of the freight industry. We will leave fighting off Godzilla to our science fiction fans.