Transportation & MobilityFebruary 26, 2024

The path to autonomous driving

Connecting transport with territory to design the cars of the future
Avatar Laurence Montanari

Advanced Driver Assistance System (ADAS) technology is central to the transformation happening across the automotive sector. It enhances the convenience and safety of every type of vehicle, from traditional internal combustion engine models to innovative electric cars and vans. Across the board, consumers have welcomed functions like adaptive cruise control, parking assistance and forward-collision warnings.

Regulators have also taken notice. In Europe, for example, all new cars must include a set of ADAS capabilities. These go beyond the more familiar functions like automatic emergency braking. Attention warnings in case of driver drowsiness or distraction, advanced blind spot monitoring to alert drivers to unseen hazards, and event data recorders also make the list.

ADAS’s purpose is to enhance the safety and security of drivers, pedestrians and cyclists. But each new ADAS technology is also a stepping-stone on the path to fully autonomous driving (AD). It is evolving so quickly that by 2030, 15% of cars sold will be fully autonomous, according to Capgemini. But developing these complex vehicles brings big challenges for carmakers and suppliers.

Defining a dynamic technology of ADAS Car

ADAS may sound like a simple concept to make drivers’ lives easier – after all, the words “driver assistance” are there in the title. But that is only part of its story. The sheer speed of innovation has already changed what the technology means to carmakers and suppliers.

To chart this dynamic journey, look at the “levels of driving automation” outlined by the Society of Automotive Engineers (SAE) International in its SAE J3016 classification. Ten years ago, the industry was still at level one. An ADAS-equipped car would have a set of cameras and sensors placed around the vehicle, which provided data for a specific task. That might be anti-lock braking or adaptive cruise control, for example, but not both.

Skip forward 10 years and carmakers have embraced level two – multiple driver support features – and made inroads into level three: conditional driving automation. Automated driving capabilities have emerged, allowing some vehicles to drive themselves in certain circumstances. Using the same sensors to feed these different functions has allowed the industry to advance ADAS technology in an affordable way. As a result, today’s ADAS vehicle is a massively complex system of software and hardware that must work seamlessly together.

More testing, less time

Speed is the single biggest concern shared with Dassault Systèmes in our discussions with auto industry customers around the world. They want to know how they can accelerate their development cycle and shorten lead times while being sure of the quality of their products and processes. These can seem like opposing goals, especially when you are developing complex ADAS vehicles.

Verification and certification present the biggest challenge in this context, because regulators need to know that the whole ADAS system will identify risks and protect other road users in any environment. The list of variables is endless: different terrains, light levels, weather conditions, traffic systems or interference from radio masts to name just a few. Proving it works in all those scenarios involves more simulation and testing than ever before.

To see how this typically plays out, look at the homologation process for a new vehicle. The carmaker has a list of requirements that must be satisfied. In physical terms, that can be simple to do. It’s easy to measure the car’s weight, for example, and that measurement will stay the same in any environment. But ADAS performance is harder to pin down.

An ADAS pedestrian detection function, for example, must be able to detect people in dazzling sunshine or blinding fog. Its sensors and camera will see things differently in cloudy, snowy or wet weather and the carmaker must prove that it will perform in each of those conditions. It takes many tests and countless kilometers of virtualization to simulate the performance of different ADAS configurations in all those scenarios – all of which add more time to the process.

It doesn’t stop there either. Every time something changes – say, a new type of headlight or traffic light is introduced – the carmaker must check and validate all those ADAS systems against the new scenario, including the ones that are already on the road. The only way to do that is through more simulation and yet more testing.

But there is a way to shorten the cycle without cutting corners.

Virtual Twin and Massive Simulation in Automotive Design

Virtual twin and massive simulation are the keys to getting complex vehicles quickly to market. When data from the same sensors is used for many different functions, vehicle design is no longer about designing individual parts and fitting them together. It’s about engineering a connected system of physical parts and computers and making sure they work together.

Today’s cars embody an intricate combination of systems. Many have more than 80 high-performance electronic control units (ECUs) onboard, all of which must work seamlessly with the physical vehicle. A system engineering approach is essential to develop something this complex.

A virtual twin of that system helps vehicle designers to make sure the whole system – physical parts, sensors, algorithms, data, chips and ECUs – works together, before going to physical prototypes and production. But ADAS adds another layer to the challenge. Each of these complex vehicles must also perform in a complex and changeable system of terrains and conditions.

Simulating the entire system of hardware, software and environment is one of the biggest, most time-consuming challenges facing ADAS vehicle developers. The only solution is to run huge numbers of simulations in parallel.

Several of our own customers are already using the Dassault Systèmes virtual twin, combined with CATIA SCANeR, to achieve this massive simulation approach. It allows them to design and test multiple models in different scenarios that are linked to real-world locations and traffic conditions. Their goal is not only to create exciting ADAS-equipped vehicles today, but also to develop tomorrow’s autonomous, self-driving cars.

Partnerships for ADAS Development Solutions

One thing that ADAS underlines is the need for a cross-disciplinary, cross-industry mindset. As a combination of systems, an ADAS vehicle is much greater than the sum of its parts. And that same principle of combining strengths has a valuable role to play in addressing the challenges of ADAS development.

Dassault Systèmes has joined forces with Capgemini to provide a comprehensive solution for carmakers to manage complexity across the entire ADAS lifecycle. It includes the tools to streamline ADAS development, accelerate market entry and reduce the cost of system engineering, simulation, verification and validation activities.

One of our customers in the auto industry illustrates this approach at work. This company wanted to speed up its development cycle by designing its own ADAS. It didn’t understand the ADAS systems its suppliers provided and even a simple modification – such as repositioning a sensor – could not be done without asking the supplier if it was possible. It was a costly, expensive process that involved a lot of extra time for validation.

But moving to designing its own systems was a challenge. The organization needed a fast, efficient way to develop ADAS itself, and to understand the technology from the beginning of the design phase to validation and certification.

A combined solution was the answer. Dassault Systèmes provided the end-to-end processes to design and validate electronic systems, hardware and software together. By partnering with Capgemini to include its leading ADAS development platform, we created a complete solution that allows the carmaker to improve its ADAS engineering and prepare for the future.

Partnerships like this are the key to helping organizations internalize the complexities as the automotive sector transitions to new technologies. ADAS involves an ecosystem of manufacturers, technology and infrastructure and increasingly, these different sectors are working together to co-develop knowledge and spur innovation.

The Future of Mobility: ADAS and the Evolution of Autonomous Vehicles

We don’t yet know what the future of mobility will look like. Experimentation continues across the world as the industry looks ahead.

ADAS straddles the automotive sector’s typical paradigms of mechanical engineering and software engineering. As it continues to evolve, it will open a debate on the future of the vehicle. Is a car a self-driving computer on wheels, for example? Or is it a physical object that people want to enjoy driving? And if carmakers continue to add more ADAS functions, how will they make these increasingly complex vehicles affordable to produce?

As carmakers explore these questions, the core focus of ADAS remains the same. Whatever the car of tomorrow can do, it must work safely in its environment. Its future lies in our ability to connect transport with territory.

Massive simulation will provide that connection. By simulating the vehicle in its environment, carmakers will find affordable routes towards safe, autonomous driving. We can expect to see a lot of engineering focused on this area, to verify that the vehicle will operate safely no matter what people, animals, traffic systems and terrains it encounters.

A new era of ADAS simulation has begun, allowing carmakers to visualize the future of autonomous vehicles and quickly validate the designs that will drive them in the right direction.

Frequently asked questions about advanced driver assistance systems (ADAS)

What is the difference between ADAS and non ADAS?

ADAS (Advanced Driver Assistance Systems) vehicles have advanced safety features like adaptive cruise control, lane departure warning, and collision avoidance systems, thanks to sensors and integration with vehicle systems. Non-ADAS vehicles lack these features and rely solely on manual operation by the driver, without such assistance or integration. ADAS vehicles are typically more expensive due to their advanced technology and safety features, whereas non-ADAS vehicles are more basic and affordable.

What is the difference between autonomy and ADAS?

The main difference between autonomy and ADAS (Advanced Driver Assistance Systems) lies in the level of control and automation in driving tasks:

Autonomy: Autonomy refers to the capability of a vehicle to operate without direct human input. Autonomous vehicles, also known as self-driving or driverless cars, are designed to navigate and operate on roads without human intervention. These vehicles rely on advanced sensors, artificial intelligence, and decision-making algorithms to perceive their environment and make driving decisions autonomously, without the need for human intervention.

ADAS (Advanced Driver Assistance Systems): ADAS, on the other hand, involves technologies that assist the human driver in the driving process rather than replacing them entirely. ADAS features include functionalities such as adaptive cruise control, lane departure warning, automatic emergency braking, and parking assistance. These systems provide various levels of assistance to the driver, enhancing safety and comfort while still requiring the driver to remain engaged and responsible for vehicle operation.

What level of autonomy is ADAS?

ADAS (Advanced Driver Assistance Systems) typically fall under Level 1 or Level 2 autonomy according to the Society of Automotive Engineers (SAE) classification:

Level 1 Autonomy: This level involves systems that provide assistance with specific tasks, such as steering or acceleration, but the driver must remain engaged and attentive at all times. Examples include adaptive cruise control and lane-keeping assistance.

Level 2 Autonomy: At this level, ADAS can control both steering and acceleration/deceleration simultaneously under certain conditions. However, the driver must still monitor the driving environment and be prepared to intervene if necessary. An example is Tesla’s Autopilot system.

Stay up to date

Receive monthly updates on content you won’t want to miss


Register here to receive a monthly update on our newest content.