Proceedings

Ramamurthy Mani, MathWorks

The development of software-defined vehicles demands processes that can help decouple software from hardware, design self-contained apps for easy updates, and automate software build and testing, while still ensuring the consistency and traceability required for software quality, safety, and security.

Model-Based Design offers distinct advantages that can help you go from software architecture conceptualization to component design and back up to software integration and testing–enabling higher quality and reliability of the final product. In this presentation, you’ll see how Simulink^® is rapidly evolving to realize this vision. Recent investments have helped create an environment that accounts for major trends in service-oriented architectures, code-based component development, virtual testing, continuous integration, and much more.

Dr. Philip Jose, Mahindra Last Mile Mobility

Electric vehicle systems are much more complex compared to IC engine-based vehicles, due to the larger number or electronically controlled power components and the expectation of seamless integration with infotainment and connectivity systems. An integrated systems engineering approach, transcending the six dimensions of integration—mechanical, electrical, signal, control, heat transfer, and mass transfer—is needed to efficiently integrate the various components in an electric vehicle so that vehicle reliability, performance targets, and program timelines are met. Rigorous breaking down of system requirements to component requirements, decoupling of component requirements, and robust test case generation to test these requirements are three key elements of an effective system engineering process. Such a process will enable different engineering teams to pursue their designs in parallel with each other while avoiding costly mistakes due to integration errors as the prototypes are built. Mathematical modeling of the system components, their requirements, and the system control strategy in an integrated manner would best serve this. However, to fully utilize the benefits of model-based systems engineering, this process must be fully integrated into the overall vehicle engineering process.

Prashant Rao, MathWorks

Learn how large language models (LLMs) work and how to build a transformer model in MATLAB^®. See a demo of an LLM-based model for MATLAB and how you can use it in your work, including which prompts to use.

Sanjeev Madhav, Tata Consultancy Services

Abhisek Roy and Sree Varshini Bhattu Saraswathi, MathWorks

The rising demand for advanced battery technologies requires frontloading of battery development. In this talk, discover how to virtualize battery development by integrating high-fidelity battery pack models, including thermal and cooling systems. Then learn how to integrate this model into a virtual vehicle and scale the simulation in the cloud. While this talk uses a battery as an example, the workflow is applicable to the development of any other EV component.

Highlights:

• Create comprehensive battery pack models with thermal and cooling considerations
• Ensure model accuracy and performance through rigorous unit testing
• Integrate battery pack models effortlessly into the virtual vehicle framework
• Leverage cloud workflows for faster, resource-intensive simulations and testing

Dr. Vivek Venkobarao, Vitesco Technologies

This presentation highlights the significance of integrating machine learning and cloud computing in the development of EV systems. During this session, we discuss:

• The rationale behind employing machine learning and cloud computing in EV system development.
• The benefits of utilizing machine learning and cloud resources in advancing EV systems.
• The hurdles and obstacles associated with the utilization of machine learning and cloud resources in the realm of EV system development.

Balasubramani Krishnamurthi, Simpson & Co. Ltd.

With electrification taking center stage in the mobility segment, the need for reliable high-performance electric powertrains with shorter development times has become even greater. Model-Based Design is a preferred approach to improving product quality and achieving a faster time to market.

In this presentation, we discuss how a system-level model helps us understand component-level requirements and interactions between the components. The team utilized insights to create component-level models using Simulink^® and Simscape™. These models were then combined to analyze system-level details and optimize performance, range, and cost.

The presentation covers the following points:

• Understanding system architecture to comprehend interactions between components and their specifications/sizes, including trade-off studies.
• Using Simulink and Simscape to model individual components such as motors and batteries.
• Integrating the component models and combining them with environmental and driver models to optimize system performance and range by changing the gear ratio.
• Testing and validating system performance.

Rahul Choudhary and Ananth Kumar Selvaraj, MathWorks

In this master class, explore the latest advancements in motor control algorithm design and deployment, specifically in the context of vehicle electrification. These algorithms play a crucial role in regulating performance characteristics such as speed and torque. The session will focus on highlighting the unique capabilities of these algorithms and providing an efficient process for their development and implementation.

Highlights:

• Motor Modeling and generating characteristics curves as a function of motor parameters
• Using MTPA/field-weakening algorithms to meet the torque and speed requirements
• Validating motor requirements and redesigning the motor based on a standard drive-cycle test
• Using Field Oriented Control Autotuner block or PID Controller block for interactive control loop gain tuning
• Generating and verifying the code with static code analysis tools
• Deploying the algorithm to the hardware and verifying rate-monotonic/control loop execution time

Prathamesh Patki, Altigreen

Over 5 lakh people lose their lives every year in India due to poor air quality from road transportation. In 2019, 21 of the 30 most polluted cities in the world were in India!

Founded in 2013 and based in Bengaluru, Altigreen designs, engineers, and produces 3-wheeled electric vehicles using proprietary and indigenously built technologies. Altigreen‘s Made in India/Made for India products are specifically designed for the environment, road conditions, and driving behaviors in India. Altigreen’s product offerings stand on four strong pillars: longest range, largest volumetric capacity, highest ground clearance, and greatest torque.

All components, including the motor, motor controller, power electronics converters, telematics/IoT, gearbox, battery, and BMS, are designed and manufactured in our 3 lakh square foot facility in Malur, Karnataka.

Altigreen uses MATLAB^® and Simulink^® products for its system level simulations and software design, and Embedded Coder^® for development. In this session, learn how using MATLAB and Simulink has helped us solve specific technical challenges related to component sizing, BMS design, and more—and reduce the time-to-market for our products.

Dr. Sandeep Anand and Abhinav Arya, Indian Institute of Technology Bombay

The Indian Institute of Technology Bombay (IITB) has established a center of excellence (CoE) for e-mobility to foster collaboration between industry and academia in addressing industry challenges. The CoE has been actively working with various companies to tackle issues concerning power electronics and e-mobility. This presentation highlights the success of the team’s simulation-driven approach to overcoming challenges related to power electronics control design and component reliability. This includes:

• A MATLAB^® simulation-based reliability study of power converters in EV applications.
• A method of anti-slip control for multi-motor single inverter EVs.
• Simulation studies on alternate capacitor technologies (electronic capacitors) for EV battery charging applications.

Dr. Jihas Khan, Chandni S Vijay, Hari Priyadharshini A., Tata Elxsi
Dr. Rishu Gupta, MathWorks

Advanced driver-assistance systems (ADAS) are an ever-evolving technology in the automotive domain that aim at improving the safety and comfort of the driver. Efficient, scalable, and diverse development and validation techniques are used to ensure that ADAS systems behave as intended. Real-time embedded systems housing ADAS applications need to be tested in a real-time environment to bring in timing and safety criticality. In this session, explore how Tata Elxsi has accomplished real-time ADAS validation using RoadRunner, MATLAB^®, and Simulink^®. The first part of the presentation showcases how camera-based ADAS features can be tested in a real-time embedded platform using the concept of frame grabber and frame generator. Over the air simulation setup and monitor camera setup will also be explained, where the entire camera ECU is brought under the scope of validation. The second part of the presentation covers how rapid control prototyping (RCP) testing can be implemented for an ACC-AEB algorithm in a real-time platform. dSPACE^® SCALEXIO and the relevant real-time compatible toolboxes from MathWorks are also used. The ACC-AEB controller algorithm and vehicle dynamics logic are deployed for real-time execution in dSPACE SCALEXIO, while the scenario and sensor simulation in Simulink along with the scenario animation are deployed in the test PC. See the benefits of using MATLAB and Simulink products to achieve real-time testing of ADAS systems.

Ananthesh Seth, APTIV
Naga Pemmaraju, MathWorks

ASAM OSI^® (Open Simulation Interface) is a generic interface based on Google’s protocol buffers for the environmental perception of automated driving functions in virtual scenarios. It is a specification for interfaces between models and components of a distributed simulation. ASAM OSI is strongly focused on the environmental perception of automated driving functions.

In this presentation, explore the capabilities of ASAM OSI in facilitating the simulation of advanced driver-assistance systems (ADAS) electronic control units (ECUs) within a virtual environment. Learn how constructing simulation models around ASAM OSI achieves a plug-and-play functionality that seamlessly integrates with any environmental simulator. This approach allows APTIV to develop a versatile platform capable of meeting the diverse requirements of customers while minimizing the need for extensive customization and enabling faster delivery. See how you can use RoadRunner as a tool that generates scenarios and creates ASAM OSI files, which are subsequently utilized for simulation and validation of the models and ECUs.

Deva Hanuma Kishore Naidu Avisineni, Bosch Global Software Technologies
Munish Raj, MathWorks

Automated driving features are a combination of multiple complex systems. Testing and validation of these features requires millions of kilometers of scenarios; however, this is not feasible to test on on-road systems. In view of different levels of automation, there is a dire need of a modular and scalable platform that allows X-in-the-loop simulation.

Having a configurable vehicle model used for software-in-the-loop testing provides substantial advantages compared to the physical test drives. It enables automated, efficient, and extensive testing methods with minimal risk and costs for time and equipment.

The current simulation framework allows you to create test scenarios with predetermined traffic behavior. However, there is a need for systematic usage of scenarios for testing and validation of automated driving systems, needed particularly in modeling the driving environment and traffic dynamics. This allows a realistic, robust, and usable environment of the test scenarios.

In this talk, learn about a cosimulation framework between RoadRunner from MathWorks and CarMaker from IPG and how to use it with a third-party simulation framework for validation. See how to bring realistic traffic models into the driving environment to ensure all complex and corner case scenarios are covered during the validation. This cosimulation platform allows you to test millions of kilometers by combining vehicle simulation with virtual simulation software environments.

Munish Raj, MathWorks

Automated driving spans a wide range of automation levels, from advanced driver assistance systems (ADAS) to fully autonomous driving. As the level of automation increases, the need for testing these features on multiple scenarios becomes important and the testing requirements increase multifold, making the need for modeling and simulation more critical. Creating virtual environments in the form of scenes and scenarios along with a testbench is important to achieve effective simulation.

In this session, learn how RoadRunner and RoadRunner Scenario™ can help you design scenarios for simulating and testing automated driving systems. See how to incorporate scenarios in a closed loop with algorithms for testing automated driving systems.

Discover how to:

• Interactively author scenarios by placing vehicles and paths, defining logic, and parameterizing scenarios
• Export and import scenario and trajectories to ASAM OpenSCENARIO^®
• Programmatically create scenario variants from seed scenarios
• Set up scenario-based validation of ADAS features like highway lane change
• Set up a test automation framework for virtual simulation

Gangadhar Malagi and Nikhil Nair, Daimler Truck Innovation Center India

The emergence of autonomous commercial vehicles has become a significant trend driven by the need to address driver shortages and enhance safety. As these vehicles heavily rely on software for their operation, it is crucial to develop robust and efficient algorithms to enable autonomous driving. This talk highlights MATLAB^® and Simulink^® products, including Stateflow^®, in the software development process for autonomous commercial vehicles.

From requirements to testing, MATLAB and Simulink provide a comprehensive platform for control and path planning software development. The complexities inherent in designing algorithms for commercial vehicles necessitate powerful tools to tackle challenges effectively. MATLAB and Simulink offer a wide range of capabilities that aid in controls and planning algorithm development, allowing engineers to address complex scenarios.

They also facilitate code deployment for prototyping, enabling engineers to validate their algorithms on real-world hardware in the commercial vehicle. This prototyping capability provides invaluable insights and opportunities for refining and optimizing algorithms in terms of functionality and performance before full-scale implementation.

By leveraging the powerful capabilities of MATLAB, engineers can effectively navigate the intricate landscape of autonomous commercial vehicles, ensuring the development of robust, efficient, and safe autonomous driving systems.

Shwetha Bhadravathi Patil, MathWorks

The automotive industry has embraced a service-based approach, known as service-oriented architectures (SOA), to design applications for software-defined vehicles (SDVs). SOA introduces a paradigm shift, emphasizing high reusability, streamlined updates, and reduced hardware dependencies in software development. It revolves around the concept of dynamic service discovery, publisher, subscriber, and runtime reconfiguration. The concept of SOA has been widely incorporated into industry standards, including AUTOSAR Adaptive, DDS, and ROS.

Join an insightful presentation on the evolutionary journey of Simulink^® in developing SOA-based applications, highlighting the following key capabilities:

• Advanced Simulink semantics for service development
• Software architecture for SOA, AUTOSAR Classic, and AUTOSAR Adaptive
• Seamless migration from traditional applications to SOA and AUTOSAR Adaptive applications

Indranil Bhattacharjee, Amazon

The future of automotive technology is cloud native and software defined. This evolution presents challenges and opportunities to leverage cloud-native capabilities from software development to safe and secure operations, as well as model-based approaches to ensure reusability, reliability, and quality. This presentation explores how AWS and its partner ecosystem’s automotive-specific solutions for connected and software-defined vehicles can be a value multiplier for automotive OEMs and their ecosystems. We discuss how model-based development and design can be integrated with cloud-native development and deployment, creating a powerful framework for developing and scaling software-defined vehicles. We also explore the benefits of leveraging cloud-native and model-based approaches, including increased efficiency, performance, agility, and innovation.

Padmavathi R, Mahindra & Mahindra
Jayanth Balaji Avanashilingam, MathWorks

Artificial intelligence is employed in a wide range of fields and applications. Learn how the use of AI has been investigated in the current effort to speed up the development of powertrains. The areas of applicability include:

1. Enhancing emission robustness by the statistical creation of a worst-case cycle and implementation for bs6.2 development.
2. Using shallow neural network techniques to create virtual sensors that can replace genuine ones for cost and maintenance advantages.

Nukul Sehgal, Rajat Arora and Vamshi Kumbham, MathWorks

In this session, explore the application development and integration of reference workflows into a continuous integration (CI) and continuous delivery (CD) pipeline, and how to streamline the software development lifecycle with a Model-Based Design approach.

The key takeaways from this master class will be:

1. Using model-based development for service-oriented architectures
2. Setting up and automating Model-Based Design verification and validation using CI/CD
3. Deploying software-defined vehicle applications
4. Setting up a test automation framework for virtual simulation

Akshay Bujone, FEV India

As the automotive industry is going through transformation with new trends and technologies, the current vehicle architecture is reaching its limits and needs to have an innovative ground-zero approach for a future vehicle architecture.

The traditional method of automotive feature development involves designing hardware and software and then integrating them into the vehicle. As advancing features are becoming more interdependent, a software-driven approach to feature development can balance and limit complexity and integration efforts. Software-driven feature development is characterized by designing and laying out central computing units able to run not only features of different domains, but also any further features to the vehicle that might be added via pure software additions after vehicle shipment.

This session highlights the FEV demonstrator project, which lays out software-defined vehicle architecture based on cross-domain applications, and to prove its feasibility by a prototype. The project uses the existing Model-Based Development workflow while developing this software-driven architecture based on the Adaptive AUTOSAR standard.

Sanjay Nimbalkar, Daimler Truck Innovation Center India

In recent years, software in the automotive industry has become an important part of success. Vehicles are increasingly becoming part of the modern, digital, and fast paced world. To meet customer demands, it is important to develop application software in service-based algorithms. Service-oriented architecture (SOA) is the solution for software-defined vehicles to enable reliable large data transfer between ECUs and for efficient usage of communication backbone via FOTA (firmware over-the-air).

This talk highlights the following topics:

• Development of SOME/IP-based models (Methods and Events)
• Automatic code generation based on SoCs
• Usage of AUTOSAR Blockset new features
• Integration of application software with AUTOSAR stack

Reena Parekh and Aditya Jain, TCS
Koustubh Shirke, MathWorks

The automotive industry is witnessing its next phase of transformation. Vehicles with electric drivetrains and automated features are becoming advanced and sophisticated with continuous over-the-air software updates. For these complex software-defined vehicles, prognostics and predictive maintenance become ever more critical than before. This presentation proposes a machine learning based framework created and deployed through MATLAB^®. It utilizes minimally labeled vehicle data and identifies and flags anomalous behavior that went undetected or got introduced with the aging of components. This framework can be adopted for large real-time or time-series data for early identification of failures and can be deployed on the cloud or vehicle edge.

Highlights include:

• Handling and preprocessing the big data.
• Developing the AI models using the MATLAB toolchain to detect anomalies in the powertrain subsystem and proactively flag them to customers or OEMs.
• Deploying the predictive maintenance solution on the cloud, ensuring scalability and accessibility from anywhere.
• Designing app-based dashboards in MATLAB with intuitive visualizations, empowering users to make informed decisions about maintenance actions.