中文 English

How Verification And Validation Can Create A Comprehensive Digital Twin To Design Tomorrow’s Self-Driving Cars Today

Creating automotive products that perform well in the stochastic and diverse real world.


By Shakeel Jeeawoody and Hieu Tran

Autonomous vehicles (AV) are the future of driving – and the future might not be so far away. Optimizing the capabilities of self-driving vehicles and the environment around them could lead to a mad dash to the financial finish line for manufacturers. Who takes the checkered flag will ultimately be determined by the efficiency and cost-effectiveness of their processes. These elements are crucial to measuring success in this highly competitive space.

This article explores how to create automotive products that perform well in the stochastic and diverse real world and can put self-driving cars on the road faster.

The Society of Automotive Engineers (SAE) defines the levels of driving automation, ranging from 0 (fully manual) to 5 (fully autonomous). Known as SAE levels, these levels have been adopted by the U.S. Department of Transportation.

As adoption of automated driving technology (SAE level 3) reaches the mainstream, industry focus shifts to the capabilities fully autonomous vehicles offer (levels 4 and 5). The fusion of vehicles with intelligent roads and the communication infrastructure is essential to the vision of a self-driving future, for several reasons.

First, autonomous cars will need to rely on road and traffic guidance from other cars and the environment to achieve safe and optimal driving. Second, driverless cars must be ready to respond to extreme scenarios in a shared environment that exceeds their training and conditioning. Finally, cars must depend on the communication and cooperation from the surrounding infrastructure for important functions such as data acquisition and processing.

These issues present challenges which, when properly addressed by the original equipment manufacturers (OEMs), open doors to new landscapes and opportunities. To tap into these opportunities, companies must work quickly to mitigate design risks, shorten the development lifecycle, and deliver products that meet or exceed industry safety standards.

As the automotive industry shifts its focus from drivetrain to software and electronics, sensor fusion and software innovation are emerging both as opportunities and challenges. To adapt and thrive, the automotive designer needs an effective environment to evaluate, develop and test automotive compute hardware solutions with a high level of fidelity, while at the same time left-shifting the software development lifecycle.

Opportunities and challenges

Electrification is disrupting the automotive industry, rearranging the supply/value chain and changing how cars in the future are architected and built. The concept of a “software-defined vehicle” opens new opportunities for existing and new players to redefine and package transportation and mobility as services. At the same time, it also presents to the automotive suppliers, manufacturers and downstream customers enormous challenges with the design and verification of connected autonomous vehicles.

The modern car is a delicate balance of mechanical, electrical, and electronic devices that are optimized for power, heat dissipation, weight, and safety, among many other factors. The software system for a high-end vehicle is estimated to contain around 100 million lines of code, and this number is increasing. Ten times more is expected for fully autonomous operations (levels 4/5). In contrast, the control system that runs a single U.S. military drone uses 3.5 million lines of code. Despite increasing complexity, the race toward autonomy and electrification, fueled by interests from data and consumer-centric companies such as Google, Apple, and Amazon, has intensified time-to-market pressure for automakers and OEMs.

Manufacturers must reconcile the need to rein in the traditionally long product lifecycle against the stringent safety, security, and data requirements of future cars. From this, the ability to design, test, validate and support the operation of autonomous vehicles in a cooperative and connected environment is proving to be a significant and ongoing engineering requirement.

Intelligent cars rely as much on external information as they do onboard sensors and cameras for safe operation. The evidence of the safety benefits from driving automation is indisputable: 94% of crashes originate from human error (National Highway Traffic Safety Association). Driver-assisted features such as left-turn assist, look around viewing, and crossing alert are integral components of the actionable artificial intelligence (AI) response loop in AVs and are paramount to safety. But autonomy is just one part of the driverless narrative. According to McKinsey, the unconstrained technology introduction of private AVs will lead to urban sprawls and a 10% increase in congestion during the next 5 to 10 years. In contrast, the adoption of shared AVs reduces congestion by 20% while adding $850 billion in annual revenue to the U.S. economy.

Cooperative mobility improves safety and efficiency, reduces emissions, and fosters growth. And connectivity and data are the keys that open the doors to these opportunities. Dedicated short-range communications (DSRC) enable wireless broadcast of traffic lights and information to extend the vehicle’s field of view and provide guidance for speed and acceleration. With fast and reliable cellular access (C-V2X), buses can broadcast routes and schedules to awaiting passengers while streaming contents to those onboard. Cars and taxis can upload real-time sensor data or send emergency requests. Special purpose AVs such as long-haul trucks and delivery vans can be remotely monitored and controlled for safety and security.

To support cooperative driving, automakers and OEMs need to customize their products for specific vertical markets, enterprises, and municipalities. And in doing so, they need the tools and environment that realistically model the communication requirements of their customers. The right environment enables automakers and OEMs to test and characterize their products under simulation without sacrificing real-world fidelity and non-determinism.

Computer vision and sensor technologies such as radar, lidar, and GPS helped lay the foundation that makes vehicles more aware of their surroundings. But automated responses founded on these technologies can be limited by unplanned conditions such as uncommon lighting angles, accidents, road work, emergency responses, etc. AI hardware learns from and detects common and known corner cases to deliver real-time control and autonomy. However, the variability in geography, environment, and regulation across diverse market segments and municipalities yields permutations of known and unknown corner cases that are combinatorically enormous.

Driverless cars must, for this reason, undergo billions of hours of training under realistic scenarios in which vehicles navigate various roadway environments potentially shared by traditional vehicles, pedestrians, cyclists, and other vehicular objects.

Software and electronics, the heart of all AV self-driving systems, present both challenges and opportunities for innovation. The AV self-driving system must tailor to the specific environment into which the vehicles deploy. Furthermore, they must continuously be updated to reflect changes in the vehicle design, mechanical and electronic components, and other external factors such as the condition and availability of roadway and communication infrastructure to meet an acceptable level of usability and accuracy. Consequently, participants in the automotive ecosystem must quickly reposition their business to adapt. Semiconductor vendors will benefit from the rise of electrification and sensor fusion in AVs, as will semiconductor infrastructure providers from the move to integrate functions into a System-on-Chip (SoC) or package (SiP) in order to reduce power consumption and improve liability and performance.

Development and verification platform

The design and verification of the connected autonomous vehicle must evolve with the shifting automotive landscape. Critical to the success of the industry are the availability of future methodologies and platforms that enable automotive architects to effectively explore different compute and sensor options in a closed-loop environment to simulate, test, and validate their design against different vehicular compute and sensor configurations, deployment options, and environmental variations. Such a methodology and platform must provide verification coverage of the complete vehicular system, enable characterization of a multi-vendor ecosystem, and benchmark different key performance indicators (KPIs) such as compute acceleration, power consumption, and costs in a continuous and closed-loop manner through flexible business models. Ideally, the platform should also be cloud-native to meet the needs of a multinational design team. The platform must be able to simulate a replica of a physical entity or asset placed in a virtual environment at a high level of fidelity.

In an automotive compute system design, the replica is the device or product undergoing development, test, or verification (device-under-x). Automotive devices-under-x extends from onboard elements (sensors, cameras, ECUs, compute solutions) to infrastructure elements (traffic lights, smart signs, wireless communication) and vehicles (cars, trucks, motorcycles). An effective development and verification platform provides the means for developers and software systems to interact with the device-under-x in a closed-loop context. Simulation helps the designers understand how an automotive product or device may ultimately function. The design and verification platform shall extend this capability to all phases of the product design and development, from conceptualization to stages of implementation and deployment.

Mixed reality
A confluence of real and virtual objects that co-exist and interact in near real-time includes physical chip-level (or on-chip level) devices, SoCs, performance accelerators and even complete vehicles together with their respective virtual models and simulations. The fusion of connected automotive elements with communication infrastructure within the platform cloud elevates the depth of details and diversity of use-cases that closely mirror real-world conditions. This enables what-if design tradeoffs to be conducted in manners that are repeatable and reproducible, facilitating evaluation and exploration.

ISA and SoCs
The instruction set architecture (ISA) defines the structure and behavior of the machine code that runs all computing devices, including electronics in vehicles. Automotive components such as ADAS SoC, CPU, GPU, DSP, FPGA, VPU, Image Processor, and NPU, along with their operating software, are profoundly dependent on the ISA chosen for them by the supplier of these components. The sensitivity of ISA to the automotive compute element cannot be overstated: cost, power, performance, reliability and security, time-to-market and competitive advantage are just a few. The mixed-reality design and verification platform must offer a detail-rich, high-fidelity environment to characterize the feasibility and impact of a specific ISA and ISA extension to a design. ISA-specific code for the device-under-x can execute inside a virtual car or an automotive infrastructure element by way of EDA tools integrated within the platform such as simulators, emulators, high-level modeling tools, and prototype boards. The given automotive element can be made to carry out certain functions, execute specific responses and generally interact with external stimulus while providing valuable KPIs such as latency and throughput as well as cycle-per instruction (CPI), speedup, threads, RAM usage and performance.

Co-design is the concept that a single language can be used to describe the platform’s hardware and software, enabling functionalities to be assigned to each as necessary. In practice, the ability to evaluate the design performance under realistic workloads is fundamental to co-design and has been difficult to achieve under simulation. Inclusion of the continuous integration and delivery (CI/CD) pipeline fast tracks performance trade-off analysis associated with the partitioning of digital components into hardware or software in co-design. Software is critical to the automotive platform. The design and verification environment enables the development of software to be left-shifted in the product lifecycle to save time and facilitate design trade-offs. Automotive companies that integrate vertical functionalities into an SoC can benefit from left-shifting to reduce time-to-market and improve quality.

Given the complex mix of mechanical, electrical, electronic, and software elements in the car, the design and verification platform provides the means for car manufacturers to evaluate the efficacy of these components from a diverse ecosystem of suppliers. Since the manufacturers likely prefer to purchase the most suitable parts for their product, suppliers can also leverage the design and verification platform’s capabilities to perform compatibility and feasibility testing. By doing so, they can offer and deliver parts to their customers that have been characterized for the highest level of integration.

Chip to cloud
The rise of data is rapidly transforming the connected, autonomous, shared, and electric (CASE) vehicle landscape. Challenges and opportunities for automotive OEMs now extend beyond onboard components to related infrastructure and services. For example, take the concept of virtual traffic lights (VTL), which work by utilizing the car’s speed, acceleration, and distance to calculate time to green. VTL can further communicate with real traffic light infrastructure to increase the driver’s visibility in poor weather and share traffic conditions and guidance with other cars. The capabilities of VTL are useful to municipalities for planning and safety purposes, while the practical benefits it provides can lead to safer road conditions and a reduction of greenhouse gas emission due to lower fuel and energy consumption.

Beyond that, the intelligent road future represents new greenfield business opportunities for carriers and edge providers. In addition, the data generated are immensely valuable to the cloud enterprises that deal in the currency of user engagement and analytics. An effective design and verification platform offers an immersive environment for suppliers and companies to explore and develop connected automotive software and services in closed-loop settings that incorporate the automotive computing device, edge infrastructure, and elements of the public or private cloud.


The design and verification of the future connected autonomous vehicle is a complex endeavor. We posited in this article a closed-loop, mixed reality, hardware/software co-simulation environment and cloud-accessible platform with high density and fidelity that can accelerate overall product development as well as shift-left software development.

Siemens PAVE360 is a complete autonomous verification and validation environment modeled as a system that represents a comprehensive digital twin image of the physical vehicle and its driving surroundings. Together with the EdgeLab.ai mixed-reality SW and HW 4/5G edge communication platform, PAVE360 powers the needs of OEMs and suppliers to explore, design, build, and train/validate compute solutions critical to the future of connected and autonomous driving cars. The combined comprehensive digital twin approach offers a near real-time, realistic, and repeatable setting needed to create automotive products that perform well in the stochastic and diverse real world.

Leveraging Siemens’ automotive ecosystem that brings a rich diversity of models across the automotive industry, these integrated solutions provide the ability to see a design from the immersive experience of the vehicles, occupants, pedestrians, motorbikes, and other objects on the road. They also provide OEMs with insight and understanding that can minimize risk, improve development time, and increase safety.

From integration and training for auto-pilot systems to the fusion of in-vehicle control/infotainment and integration with data and services of the automotive edge and cloud, the platform offered by PAVE360 and EdgeLab will prove to be a crucial component of any automotive company’s process and development portfolio. We believe that this powerful combination is the only available comprehensive and significant platform that is deployment ready in the industry today.

Shakeel Jeeawoody is the Strategic Alliances product marketing manager at Siemens EDA.

Hieu Tran is the president and CTO of Viosoft.

Leave a Reply

(Note: This name will be displayed publicly)