Proceedings

Christian Dziobek, Dr. Thomas Ringler, and Dr. Florian Wohlgemuth,
Mercedes-Benz

Explore how your longstanding journey with Model-Based Design can help Mercedes-Benz to tackle future challenges.

Highlights:

• The first production-ready, auto-code-generated software functions developed with Model-Based Design in Mercedes-Benz series cars in the early 2000s
• Initial ideas and vision of virtual distributed system simulations
• Roadmap towards an integrated tool chain based on an automotive standard
• Implementation steps
• Roadmap and results
• Success factors

Tunc Simsek, MathWorks

Our driving experience will soon be defined by the software running in the car and in the cloud. A new holistic approach for software architectures based on services and service-oriented communication is emerging. This approach enables continuous development and deployment of innovative software features and makes new types of collaboration possible between OEMs and software platform providers. Learn how the Simulink^® product family is evolving to be a technology accelerator for software-defined vehicles. Discover new capabilities to model, simulate, and deploy models to modern service-oriented architectures.

Angel Martinez Bernal, Real-Time Innovations

Simulink^® enables users to model and simulate different time-dependent systems before going into production. To accelerate the development cycle, automotive engineers are increasingly incorporating real-world data streams directly into models from the earliest stages. Data Distribution Service (DDS) is a publish/subscribe middleware used in production-grade automotive distributed systems that provides the data backbone for systems to be accurately modeled with the same inputs and outputs from real-world conditions through Quality of Service (QoS) configuration. DDS Blockset provides apps and blocks for modeling and simulating software applications. DDS blocks can be added to Simulink models, enabling the connection to other components through the DDS communication framework. This allows for the introduction of QoS into the earliest phase, to reduce project risks and costs of system development. Get an introduction to DDS and learn how to work with DDS Blockset to simulate and implement real-time communications between Simulink applications and other DDS-based applications.

Shigeki Hasegawa, Kyoto University and
Toyota Motor Corporation

The worldwide investment to businesses involving carbon neutrality and decarbonization has become a priority after the Kyoto Protocol in 1997 and the Paris Agreement in 2015 for the prevention of global warming. Hydrogen (H2) energy is considered one of the most promising alternative energy sources to fossil fuels from the viewpoints of storability, portability, and productivity from a variety of energy sources such as solar power and wind power.

The fuel cell (FC) system plays a main role in the enhancement of H2 utilization due to high efficiency compared to conventional internal combustion engines. FC system manufacturers demand the development of a variety of system products for automotive, stationary, railway, marine, and aviation purposes. Despite such an expectation, development of FC system products is conducted by trial and error because the behaviors of the FC system are highly complex. Significant cost and efforts are required for manufacturing the prototypes, calibrating the controllers, and testing the system responses. It is one of the highest barriers for entry to the FC industry and enhancement of H2 utilization. From such backgrounds, implementation of model-based development to the entire FC system development process is demanded by the FC industry. In Model-Based Design, the specification of the system components and controllers can be determined by considering the interactions among them before manufacturing the prototypes.

Though studies of the entire FC system are important, the fuel cell itself has been more intensively investigated. Less research has been done on the FC system including the FC stack; the system components of air, H2, and cooling subsystems; and the FC system controllers. In addition, a system simulator that can estimate the dynamic behavior of the FC system—and can take the interactions among the FC stack, FC system components, and controllers into account with an acceptable computational speed and accuracy—has not been proposed yet. The objective of this study is to develop the integrated system simulator, including the FC system component models and controllers in the entire FC system, which can be utilized as a design platform for the wide range of applications such as automotive, railway, marine, aviation, and stationary power generation purposes. The authors have developed the one-dimensional (1D) FC system model including the 1D physical models of the FC stack in previous research and the FC system components of air, H2, and cooling subsystems. The developed simulator can estimate the dynamic behavior of the entire FC system with acceptable computational speed and accuracy. It is possible to modify the simulator according to the system requirements of the various applications.

Dr. Marco Gecić and Dineshkumar Selvaraj,
Infineon Technologies

See a practical case study showing that a virtual software development environment and testing methodology, free from the constraints of hardware lab setups, can accelerate development time.

The key in this environment is the need to “shift-left” development activities. This shift becomes important as performance and efficiency are challenging the OEMs, pushing the need to make decisions early in the design cycle. Learn about the need for a virtual hardware-based Software Development Kit that can allow early algorithm studies of the control system from concept through to binary optimized control software implementation.

Where the automotive ‘V’-cycle traditionally uses simulation at the early stages of product definition with MIL and moving to SIL, the HIL stage has many dependencies and access issues. It is necessary to have a more complete software development kit on a virtual target early in the design cycle to accelerate development and ensure quality before reaching the board. Through a practical EV motor control case study, discover the concept of a virtual hardware-in-the-loop (vHIL) SDK environment, which allows a Synopsys^® virtual prototype of the Infineon^® TC4xx MCU to be coupled with other simulators (in this case Simulink^®) to form a fully virtual, high visibility closed loop simulation with plant models. It also allows a highly efficient environment with which to develop, test, and verify binary target software that is ‘board-ready.’

Dr. Irina Kaiser, Bosch Engineering

Bosch Engineering presents an ADAS platform that allows test activities, first calibration of features participated on different ECUs, and includes an automated tool chain for scenario execution and evaluation of measurements. This talk also shows:

• Radar-based ADAS features and virtualization of Bosch ECUs
• A customer-specific solution for SIL tests as a virtual platform on a vehicle simulator
• Integration of many Bosch components to create a complete ADAS system in a virtual environment

Dr. Stephan Kirstein and
Dr. Roxana Daniela Florescu, Continental

There are a large variety of AI learning frameworks. If you are interested in a particular convolutional neural network, you are restricted to the framework it was originally developed in. Often Docker containers are used to run different networks at the same computing hardware. When running different networks into a test vehicle, a standardized way of deployment is mandatory instead of maintaining different Docker containers with competing requirements to the GPU driver and libraries. It is a comfortable handling of the complete vision stack, including image acquisition, network inference, and all preprocessing and postprocessing steps.

MATLAB^® and Simulink^® provide many image processing functions and supports to run neural networks based on the Open Neural Network Exchange (ONNX) format—an established standard in the community. Furthermore, the capability of C/C++ code generation is beneficial for in-vehicle usage.

In this presentation, see different deployment options using CPUs, GPUs, standard PCs, or embedded devices.

Arjo van der Ham, Lightyear

The three main concerns for buying EVs are range, cost, and charging. Energy consumption is the key driver for all three, and therefore is the main focus of Lightyear’s mission to deliver clean mobility for everyone.

To achieve roughly half of the energy consumption of current EVs, Lightyear has developed the Lightyear 0 solar car from the ground up, including 4 motors directly in the wheels and 5 square meters of solar panels. Besides an overview of where the company stands today, how they got there, and a peek into where they're going, Lightyear’s CTO dives into the Model-Based Design process that enabled this radical reduction in energy consumption, the technology that resulted from it, and how MATLAB and Simulink are used along the way.

Stefan Teuchert, MAN Truck & Bus

MAN develops commercial vehicles that are—thanks to digitalization—fully connected software and server platforms. In this keynote, learn about the possibilities of the current systems and how the development process is enabling those possibilities. New boundary conditions are leading to a new development approach. Hear about different perspectives of the approach and the roles of Android Automotive, ROS, and Adaptive AUTOSAR. MAN is in the last step of transformation towards becoming a software company.

Penny Anderson, MathWorks

Transforming current automotive business models requires vehicles to be software-defined and a fresh take at how that software is built. Model-Based Design is already being leveraged with containers and continuous integration pipelines to design, test, and deploy software-defined systems on modern vehicles. Hear how MathWorks is partnering with the industry to extend its offerings—making them more turnkey to support methods and process in DevOps.

Dr. Tjorben Gross, MathWorks
Skanda Naglapur Ramamurthy, MathWorks

Faster delivery needs and rising software complexity are common trends in automotive organizations. This raises numerous challenges for conventional software development processes. Questions to be answered during this transformation include:

• How can agile teams develop a comprehensive issue detection process?
• How can software architecture and quality be improved while still adhering to industry standards like ISO 26262?
• How can advanced workflow options be leveraged with cloud platforms?

In order to answer the questions above and to improve DevOps metrics, you’ll learn about best practices for interactive and automated development during this presentation. Best practices include continuous integration workflows and a shift-left approach using Model-Based Design and Polyspace^® in your software factory. You’ll also discover methods for lowering the risk of delayed software delivery and increasing confidence in software quality.

Dr. Hugo de Kock, MathWorks
Dr. Jan Janse van Rensburg, MathWorks

Companies want to make more use of virtual development to reduce time-to-market and development costs. Building physical vehicle prototypes and testing them under different climatic conditions is expensive, and it takes a long time for prototypes to become available. Virtual vehicles offer the possibility to start development and verification/validation much earlier ("shift left"). However, some of the technical challenges are to integrate many plant models from different physical domains as well as software models, to select appropriate fidelity for the required analysis, and to make the models and tools available to a wide community. MathWorks offers solutions to overcome these challenges, allowing you to get started quickly, continue working on a trusted and shared platform, be flexible, and scale via the cloud. Learn how to perform some of the key workflows for building an electric virtual vehicle and moving from desktop studies to large-scale studies in the cloud.

Sriram Mandayam, Volvo Cars

Learn how Volvo Cars designed a robust modular simulations platform for simulation, analysis, and optimization of engineering systems with dynamic control blocks. The systems are modeled to accommodate multifidelity models based on the depth of physics and the purpose of application. Based on the requirements, one can use a 2D or a 3D model within the same architecture at the expense of time for better resolution in specific parts of the systems. The components in the system are modeled for continuous development (CD) and continuous integration (CI). The system is composed of vehicle dynamics, transmission, electrical, and thermal systems. The systems are modeled using the conservation principles of classical physics.

Creating these models and systems in this fashion helps in estimating energy efficiency, performance, and design validation in all stages of the vehicle development process. The systems and modular approach allow for a single platform used for multiple dedicated purposes, such as SIL and HIL testing. The solution execution times are so small that a large combination of vehicle configurations can be simulated—reducing the CAE cost, energy usage, and CO2 emission and generating data. In the future, the collected data will be used with machine learning to improve efficiency in real-time drive.

Katja Deuschl, Mercedes-Benz

Advanced technologies such as artificial intelligence offer new opportunities to improve existing software development processes in a modern vehicle. Oftentimes such improvements can be accomplished through exact knowledge of critical vehicle states and inputs. Using physical sensors for such tasks can be expensive or even impossible, and the implementation of an alternative virtual sensor using artificial intelligence offers significant advantages. However, many times the deployment to embedded hardware can be challenging. Typically, memory footprint is crucial on embedded hardware and production code needs to be optimized. In this session, see how a virtual pressure sensor, developed with a recurrent neural network, can be implemented in a production code generation tool chain using only fixed-point datatype. The newly implemented workflows are fully automated and were developed in a joint project between MathWorks and Mercedes-Benz.

Gernot Schraberger, MathWorks
Lorenzo Nicoletti, MathWorks

See how a battery electric vehicle model and drive cycles can be used to optimize the topology and the number of cells in a battery pack to fulfill targets like fuel economy, range, acceleration, and costs.

Simone Hämmerle, MathWorks
Dimitri Hamidi, MathWorks

Building automated driving systems is a complex task that spans multiple disciplines. Discover new features and examples in R2022a and R2022b that will allow you to:

• Create scenes and scenarios for driving simulation
• Simulate sensors for automated driving applications
• Design planning, control, and detection and tracking algorithms
• Deploy to C, C++, GPU, and ROS

Dr. Gaspar Gil-Gomez, MathWorks
Dr. Mohammad Abu-Alqumsan, MathWorks

The automotive industry has reached a clear consensus that virtual simulations are crucial to validate the safety of the intended function (SOTIF). Explore a conceptual framework that guides these virtual simulations based on ISO 21448.

Advait Valluri, MathWorks

Learn how to design a controller for vehicle platooning applications with vehicle-to-vehicle (V2V) communication. Every following vehicle in a platoon maintains a constant spacing from its preceding vehicle. Vehicles traveling in tightly spaced platoons can improve traffic flow, safety, and fuel economy. Each vehicle obtains the position and movement information of the other vehicles in the platoon wirelessly via the V2V communication. A given acceleration profile drives the lead vehicle, and every trailing vehicle follows the lead vehicle while maintaining a predefined space by a platooning controller.

Maxime Francois, MathWorks
Simone Hämmerle, MathWorks

The development of reliable automated driving software depends on the test environment. During software development, a large number of corner cases need to be tested and additional cases have to be discovered. For this reason, the use of simulation is vital to perform a rigorous, controlled, and extensive testing of the vehicle operating conditions.

In this master class, see how you can use MATLAB^®, Simulink^®, and RoadRunner to design and simulate realistic driving scenarios and explore:

• Designing scenarios with Automated Driving Toolbox™
• Building 3D scenes with RoadRunner
• Designing scenarios interactively using RoadRunner Scenario
• Creating RoadRunner scenarios variations programatically
• Performing co-simulation of RoadRunner with Simulink