Autonomous driving vehicles collect and process more than 40 terabytes of data in a single day.
By Robert Knorr, Marketing Lead, Bressner
The idea and dream of a fully autonomous car is almost 100 years old. It was first formulated in the US magazine Scientific American. The reason for this was presumably the "American Wonder", a remote-controlled car that drove through Manhattan in 1925. After almost a century, it seems that the automotive industry is on the verge of reaching its goal of having driverless and fully autonomous vehicles participate in everyday traffic. But when will it finally come? To answer this question, we need to examine the reasons why vehicles are not yet fully autonomous.
What is autonomous driving?
The Society of Automotive Engineers has defined five basic classification levels for automated driving. Level 0 is marked as "no automation" and level 5 as "full automation". Vehicles equipped with level 3 automation systems have been available since 2016, and levels 4 and 5 are now increasingly being used in vehicles for testing and development purposes.
Level 1: Assisted driving
The driver has full control of the vehicle and is supported at most by individual assistance systems such as cruise control.
Level 2: Partially automated driving
The vehicle takes control within certain lateral movements, for example when overtaking or parking.
Level 3: Highly automated
Under certain conditions (manufacturer-dependent), the vehicle takes over the steering, braking and accelerating. The driver must be able to react immediately to signals from the system.
Level 4: Fully automated driving
The vehicle handles all tasks completely independently. No driver attention is required, as the vehicle always maintains a safe and compliant state.
Level 5: Autonomous driving
The vehicle drives completely autonomously, so that not even a driver is needed. Steering wheel and pedals are not needed.
Status Quo
The U.S. autonomous vehicle market was estimated at over USD 3 billion in 2020. By 2026, this is expected to grow to over USD 8 billion (26% CAGR). A similar picture can be seen in the autonomous trucking sector where the market is expected to grow from USD 460 million to 1.55 billion (CAGR 22.4%) in the next few years.
Currently, the US market for autonomous vehicles is still dominated by conditional automation (Level 3). However, there are still some limitations in Level 3, such as the driver's constant attention. For this reason, more and more car manufacturers are pushing directly toward vehicles with Level 4 autonomy. This development is already noticeable in the field of autonomous trucking, among others. Here, the focus is clearly on the development of trucks with automation levels 4 and 5. Self-driving truck developer Torc Robotics for example is already testing its Level 4 technology not only on open highway, but also in more challenging environments where the ai must navigate intersections with traffic lights and share the road with other road users.
Challenges in the development of autonomous vehicles
Images of autonomous driving vehicles are appearing more and more frequently in the news - but at the same time, news of crashed autonomous vehicles is also increasing. This clearly shows that the technology is not yet as advanced as it actually seems.
The greatest technical challenge in the development of autonomous vehicles lies in the correct assessment of the systems and the associated decisions that the systems should make. This is due, among other things, to the huge amounts of data that have to be processed and used during the journeys. The German car manufacturer BMW determined during test drives that an autonomously driving vehicle requires approximately 40 terabytes of storage space to be able to store and process all traffic information during the day. Artificial intelligence must learn what humans have acquired through experience - because misjudging situations can be life-threatening on the road. Even though level 4 autonomous vehicles already function confidently on certain test routes today, and test vehicles already perceive other road users or cars with a high degree of reliability thanks to technological progress, this flood of data and the associated problem of making correct decisions is the biggest technical problem. To overcome this challenge, we need either hardware that is powerful enough to apply AI at the edge.
|
Regulation for autonomous driving As mentioned earlier, one of the main difficulties is getting the autonomous vehicles smart enough to make the right decisions consistently. But regulation around autonomous driving is also a complicated issue. Who is liable if something happens? What is allowed and what is not? Which roads are the vehicles allowed to drive on? In general, technical developments have already outpaced the introduction of laws. Reuters Research published an overview of how far the regulations in the U.S. states already is. California currently tops the list with the most extensive regulations, allowing the testing and deployment of robotaxis, but not autonomous trucks. Other states are also very advanced with their regulations, but there is no clear and uniform structure, which is urgently needed. For the reasons mentioned above, it is not possible to make a reliable prediction as to when vehicles will be able to participate fully autonomous in road traffic. Regardless of whether they are used commercially or privately, companies that have committed themselves to autonomous driving face the great challenge of developing vehicles so intelligently that they no longer make mistakes. Here, technological progress must initiate the next and possibly final stages before the goal is reached. In addition, legislators face the challenge of creating clear regulations that allow autonomous driving not only in certain areas or under test conditions. Only when these challenges have been eliminated will there be nothing standing in the way of achieving the goal. |
 Reuters-Research-Regulations: Autonomous driving regulations in U.S. states; as of June 16, 2022. |
Click the buttons below to share this blog post!
The character of modern warfare is being reshaped by data. Sensors, autonomy, electronic warfare, and AI-driven decision systems are now decisive advantages, but only if compute power can be deployed fast enough and close enough to the fight. This reality sits at the center of recent guidance from the Trump administration and Secretary of War Pete Hegseth, who has repeatedly emphasized that “speed wins; speed dominates” and that advanced compute must move “from the data center to the battlefield.”
OSS specializes in taking the latest commercial GPU, FPGA, NIC, and NVMe technologies, the same acceleration platforms driving hyperscale data centers, and delivering them in rugged, deployable systems purpose-built for U.S. military platforms. At a moment when the Department of War is prioritizing speed, adaptability, and commercial technology insertion, OSS sits at the intersection of performance, ruggedization, and rapid deployment.
Maritime dominance has long been a foundation of U.S. national security and allied stability. Control of the seas enables freedom of navigation, power projection, deterrence, and protection of global trade routes. As the maritime battlespace becomes increasingly contested, congested, and data-driven, dominance is no longer defined solely by the number of ships or missiles, but by the ability to sense, decide, and act faster than adversaries. Rugged High Performance Edge Compute (HPeC) solutions have become a decisive enabler of this advantage.
At the same time, senior Department of War leadership—including directives from the Secretary of War—has made clear that maintaining superiority requires rapid integration of advanced commercial technology into military platforms at the speed of need. Traditional acquisition timelines measured in years are no longer compatible with the pace of technological change or modern threats. Rugged HPeC solutions from One Stop Systems (OSS) directly addresses this challenge.
Initial design and prototype order valued at approximately $1.2 million
Integration of OSS hardware into prime contractor system further validates OSS capabilities for next-generation 360-degree vision and sensor processing solutions
ESCONDIDO, Calif., Jan. 07, 2026 (GLOBE NEWSWIRE) -- One Stop Systems, Inc. (OSS or the Company) (Nasdaq: OSS), a leader in rugged Enterprise Class compute for artificial intelligence (AI), machine learning (ML) and sensor processing at the edge, today announced it has received an approximately $1.2 million pre-production order from a new U.S. defense prime contractor for the design, development, and delivery of ruggedized integrated compute and visualization systems for U.S. Army combat vehicles.
Get the latest news, product releases, upcoming events, blog posts and more!
