Prototype to Production: Accelerate EV Development
Accelerate electric vehicle (EV) development from prototype to production using Model-Based Design.
Published: 8 Aug 2021
Hello, my name is Govind Malleichervu I work in the automotive industry segment at MathWorks. Today, I want to talk about how you can accelerate electric vehicle development from prototype to production using model-based design. Developing electric vehicles or electrified systems broadly covers systems software and data disciplines. People in these areas face many challenges. Let's look at a few of them today.
If the system optimized for maximum range, for example, is the battery size appropriate? Software reuse for example, how can I reuse the state of charge estimator from the first prototype that uses a fixed point processor, in the second prototype that uses a floating point processor? How can I quickly test prototype algorithms? Sticking to state of charge, how can I test the state of charge estimator with a battery model? How to efficiently use data and development and operation phases, for example, how can I combine data from accelerated aging and the vehicle fleet to predict battery remaining useful life? And finally, how to meet industry standards such as ISO 26262.
MathWorks provides tools to address challenges across these multiple disciplines to develop electric vehicles or electrified systems. With reference examples for electric and hybrid vehicles, you can perform design trade-off analysis to assess optimum battery size for maximum range. You can develop models rather than code, as one customer said recently, the model can serve as a single source of truth. You can automatically generate code from models for your chosen target, and this is essentially model-based design.
You can leverage data to create data driven models for electric vehicle system components. You can explore data to extract events of interest and build models, either statistical or AI. And finally, MathWorks has tools certified by TÜV SÜD as qualified tools according to ISO 26262 for ASIL A through B. Digging deeper, there are different pathways for electrification. It's clear that there are a lot of components, subsystems within the system with variants. Irrespective of the electrification technology you choose, you can benefit from model-based design. Irrespective of the system or the component that you're developing, we see that the product development process looks similar.
You analyze data. You develop models of physical systems using data-driven models or using first principles. You can simulate these models, and you can design algorithms to control the system. And in a virtual world, you can combine the physical system model, our plant model, with a controlled system model. This is also known as model-in-the-loop testing. This is really relevant in a work from home environment, where access to hardware or labs is difficult.
In addition, you don't want to damage any expensive prototype hardware. Once your control system algorithms are refined, you can automatically generate code targeted for an electronic control unit. You can then move on to testing, starting with software-in-the-loop testing, and then hardware-in-the-loop testing, before testing in the real world. And this process repeats throughout the product lifecycle, even after the product launch.
With model-based design, you have a single workflow instead of many disjointed workflows, and you can use simulation to verify and validate the system design, architecture, interfaces between components, and the components themselves at the unit level. And you can automate each of these steps. By the way, software development, as a part of this process flow, is compliant with ISO 26262.
Why MathWorks? MathWorks enables front loading of development through a systematic use of data and models. MathWorks provides tools for developing components, subsystems, and systems that make up EV's, and MathWorks has proven experience in the EV space, covering the full electrified powertrain system, our subsystems, or just the components. Some established companies today were startups in the not so distant past. And just to name a few, Tesla is a great example. Tesla used model-based design to build the Tesla Roadster, and in this article, they mention about continued refinement of MATLAB models on energy flow.
This enabled them to tune hardware, to increase efficiency, and consequently the range. Nuvera is another example. They use model-based design to develop and test fuel cell systems in a virtual environment before going to the hardware. And they were required by the Hyster-Yale group. And the fuel cell systems they developed are now used in Hyster-Yale's forklifts.
More recently, Romeo Power, a battery maker developing batteries along with battery management systems, noted that modeling and simulation with MathWorks tools is faster, safer, and less costly than building physical prototypes. MathWorks has worked with people in many areas of new product development.
We have heard them say that these are the most relevant areas, so which area is most relevant to you and now? Contact a MathWorks sales engineer specializing in electrification if you have any questions, or if you would like to attend a webinar. You can also learn more by visiting the solutions pages on automotive or electronics control design and power system analysis and design. Thank you.