The automotive and transportation industry has come a long way from the Benz Patent Motorcar. Car ownership has evolved from a hobby of the wealthy to a near-necessity of modern life. And the cars have changed as well, from the utilitarian Ford Model T to the highly customizable and feature-rich vehicles of today. The Model T was a mechanical marvel in its simplicity and was output faster than any motorized vehicle had been before. But this came with some compromises – there was only one model and it only came in black. Since then, a mix of customer demands, intensified competition, and regulatory requirements have driven automakers to produce vehicles with more horsepower, better fuel efficiency, more cupholders, and exciting new features. Today, these needs are pushing original equipment manufacturers (OEMs) beyond the mechanical realm. Connectivity features, electronics, and software are forcing the digital transformation in today’s automotive market and, as a result, many traditionally mechanical or hydraulic systems are being replaced by electronics and software.
The divergence of automotive value from the mechanical domain to software and electronic systems has upended the traditional methods for vehicle development, and not just in the design tools. Mechanical systems are actuated by electronics that are controlled by software. Ensuring these integrations are both accurate and reliable has become a critical facet of modern vehicle development.
Previous methods for orchestration won’t hold up to the rigors of modern vehicle development. Now, more stakeholders from a wider array of engineering domains must work simultaneously on systems development and integration. The document-based approach of early systems engineering efforts has been increasingly strained by the level of interconnectedness in products. The complexity of both modern and future automobiles and transportation systems will require digitalization and a solution built on the principles of model-based systems engineering (MBSE). A modern MBSE solution will enable automotive and transportation companies to manage the complexity of advanced features, increasing cross-domain integration, and growing external pressures, helping companies to build successful programs.
More features, more complexity
The start of any vehicle program is the definition of goals, all parameters, and target requirements. Will this product be a city commuter vehicle, prioritizing fuel efficiency and maneuverability or will it be used to haul large loads on a regular basis? And what features are customers likely to demand? Combining these types of questions with vehicle requirements such as the powertrain type or the efficiency target constitute a product definition. In MBSE, this is applied to the digital twin and is shared among all stakeholders. This definition, or architecture, is expanded further to include the necessary features in the vehicle – does it need an infotainment system? What driver assistance features will be included in this model?
These features are the new selling points for most vehicles, it’s no longer the horsepower rating of the engine or whether it’s an automatic gearbox or a manual one. This shift in features is compounding the complexity of a modern vehicle’s architecture. Digital dashboard instruments require sensors for each of the values being displayed, along with a control unit to log, translate, and transmit the information to the dash where it then needs to properly display the information. This seemingly simple feature requires integrating the mechanical systems being tracked, the electronic systems processing the data, the electrical networks that transport the signals, and the software to ensure each element works in tandem to deliver accurate information to the driver in a timely manner.
And this problem of complexity grows with more integrated features. Autonomous driving systems will essentially integrate the entire vehicle into a single system – the drivetrain, the safety systems, the entertainment suite, and much more – ensuring they’ll be the most complex vehicles ever produced. And instead of working on these different subsystems in isolation until the integration phase, MBSE provides a single environment for engineers as they work from the system architecture. Each engineering team can work to deliver the requirements specific to their systems or components, and ultimately produce specific outputs that are compatible with the other vehicle systems. This approach ensures internal teams and suppliers can deliver working components, and frequently accompanying software, that integrates seamlessly with the rest of the system.
No more isolation
A critical benefit of a robust MBSE approach is its ability to remove the silos that have long persisted in vehicle development. Take an emergency braking feature for example. The basic goal of such a system is to engage the brake when an obstacle is detected in the vehicle’s immediate path. To enable such a system requires sensor data about the obstacle, information on the current dynamics of the vehicle, and a method to send the stop signal to the braking system to apply brakes quickly and safely. Each of these steps involves integrating a new vehicle subsystem or component to enable the automated braking feature. Managing these connections across vehicle systems and engineering domains through traditional workflows takes time and provides opportunities for miscommunication, which can lead to suboptimal designs or, in the worst-case, failure.
Instead, MBSE ensures everyone remains in the loop as the definition is refined by the different teams. This single source of truth not only saves time and improves accuracy in communicating design data, it enables reuse of previous work. The emergency braking system for example, could leverage braking information already being employed by an adaptive braking system so decelerating the vehicle doesn’t result in a loss of control. Or the vehicle dynamics information may already be tracked by another system and can be pulled into the emergency braking feature to save time in development.
Understanding these system interactions is also critical for the development and deployment of software in new vehicles. A single OEM can have multiple models of a vehicle going out in a single year, but within each model there are likely hundreds of variations to the design – it could be different feature packages or even different suppliers for a single part. No matter what the exact make-up of the vehicle is, the embedded software managing each vehicle feature and function must be compatible and portable with every possible vehicle configuration. In the age of over-the-air updates, this challenge extends beyond the production line to vehicles in the field. OEMs will need to keep track of each vehicle and its related software configuration to verify, validate, and deploy software updates that are compatible with each customer’s vehicle.
Optimization and validation
Electric vehicles (EVs) and autonomous vehicles (AVs) face many new development challenges compared to traditional combustion vehicles. EVs have a tighter balance between range, weight, and aerodynamics due to a less energy-dense storage medium. AVs will require more coding and software testing than before to reach level 5 autonomy – or full autonomy without need for driver interaction. The solution for many of these hurdles is simulation, maybe computational fluid dynamics testing to determine wind resistance of the body, or simulated scenarios for the autonomous vehicle control system to train many more miles without the time penalty. But integration of these systems makes testing their interplay difficult if not impossible.
MBSE provides a framework for the optimization and validation of new vehicle architectures. The range of an EV can be balanced against weight and aerodynamics while also understanding the impact to vehicle dynamics. An AV system can be validated more rapidly with more comprehensive simulation tools. Artificial intelligence (AI) can even be leveraged within the MBSE methodology to progressively alter the virtual situation an AV system encounters to grow the number of variations in driving situations to optimize the system’s robustness in making driving decisions. And the development cycle of many vehicles extending beyond the production floor, data picked up from car sensors already on the road, can be used to further refine driving models. The vehicles become a continuously updated product, adding improved safety and possibly additional features compared to how they rolled off the assembly line.
MBSE for now and the future
The complexity of vehicles is increasing, even without the addition of mass electrification and automation. Fortunately, MBSE is designed to orchestrate that complexity. MBSE enables engineering teams and organizations to track cross-domain vehicle requirements from their initial definitions through to implementation in vehicles coming off the production line. Such a holistic view of the vehicle development process also offers avenues for more frequent and effective collaboration within the OEM and throughout its supply chain. Through collaboration and a definitive, single source of truth, automotive OEMs and suppliers will be able to bring innovative, exciting, and high-quality vehicles to market faster and more reliably.
The cars we drive have evolved from purely mechanical innovations with limited variation into a highly personalized and multi-disciplinary feat of engineering. Document-based methods cannot handle the complexity of these modern vehicle programs, resulting in unacceptable development timelines and the introduction of potentially crippling errors. MBSE is the next step in integrated product development and can effectively coordinate across multiple engineering domains and global supply chains with the digital twin to enable the advanced vehicles of tomorrow.