The Skyrocketing Role of Software in Cars Affects Automotive Supply Chains

The amount of software code going into automobiles has reached to 40% of a vehicle's design.

With gas prices reaching recent highs of more than $4 per gallon, it's no surprise the Chevy Volt electric car has debuted as a white-hot product, sold out at many dealers not just for this year, but for 2012 models as well. The Volt is the right car for the right time.

Behind the scenes is the story of how General Motors was able to get the Volt to market, but in many ways, it's a story of how the growing role of software code in cars has helped transform the automotive supply chain.

Opening up a new market is a huge bet at any time. But GM made a commitment to the Volt, which involved a unique pairing of a traditional gas-combustion engine with a lithium ion battery capable of driving up to 40 miles on electric power.

GM couldn't afford to miss its promised deadline of June, 2011 in getting this high-profile car to market. But it also couldn't afford to cut corners on quality, disappoint customers or lose an early lead in the next-generation electric vehicle market.

To meet the challenge, GM worked with IBM to meld software development tools with a systems design approach, enabling the automaker to develop the Volt in an unprecedented 29 months and roll out a car that has been a market success.

GM's work behind the scenes is an unmistakable indicator of where auto development is headed. Increasingly, a crucial part of the success of car development depends on software quality, not just the quality of the software systems onboard the car itself, but also the quality of the software-driven engineering processes that produce the cars.

GM had to tackle two issues that are becoming the vanguard today's car production. First, automakers need to get to market quickly and efficiently managing an increasingly complex and far-flung design and production process. Many teams work on cars together using many different tools and many loosely coupled parallel processes.

Secondly, cars have become massive software puzzles. The amount of software code going into automobiles has skyrocketed to 40% of a vehicle's design. The Volt, with its 10 million lines of code -- or about four million lines of code more than an F-35 fighter jet -- is a classic example of this trend.

Despite the incredible amount of software, car production can't be about tackling a computing problem. It has to be about building a smart set of systems, across the supply chain-- systems, which are so interconnected and interoperable that they unleash innovation and productivity gains.

The first step GM took toward making this vision a reality was to borrow some of the tools, processes and methodologies used in systems engineering and embed them into its design process. The result has been a single collaboration process that allowed GM to put in place teams around the world, which worked on the car nearly 24-hours-a-day using the same synchronized tools and systems.

This collaboration made it easier to automatically keep the design team and suppliers in the loop on changes. When changes were made during the design and manufacturing process, electronic notifications were instantly distributed. Meantime, everyone working on the Volt had a dashboard view that provided updates of the car's development. These changes all helped improve efficiency enormously and helped to get the Volt launched on time.

The potential for this kind of collaboration, and the innovation it inspires, is enormous. At IBM, we recently began working with Germany's INCHRON, which provides real-time software for embedded systems and networks, on solutions that will let developers of automotive systems model and test the behavior of motor control components and systems. This process allows suppliers to model and test different design approaches based upon the latest information from their automobile manufacturing partners. As a result, timing and performance requirements can be understood earlier in the design process. The software helps cut development time and costs by up to two-thirds.

As part of the Volt's production, GM tackled another issue in the design and manufacturing process: reducing complexity. Embedded software systems aren't just becoming a bigger piece of the cars we drive, they're becoming increasingly complicated. This rising complexity is eating into efficiency and productivity at a time when everyone in the automobile industry is faced with having to do more with less. So a key goal of the GM team was to pare back the complexity of these embedded systems and to try reuse software code across various lines of vehicles through strategic and smart software development.

The process also demonstrated that sophisticated software tools and supercomputers can be put to work in a myriad of ways. They helped model the interactions of the Volt's embedded systems with the electric-drive systems, putting the 161 components that make up the Volt's battery through validation tests for short circuits, dust, corrosion, impact, water submersion, as well as crash and penetration tests. The supercomputers were also used to run crash-test simulations, cutting down the number of physical crash tests to less than 400 per year from an average of 600 three years ago, as a result, saving tens of millions of dollars.

Today's economy is now irrevocably based upon sharing information. Markets are much less protected. Brand loyalty has to be earned with every sale. That's why connecting together the systems that underpin the automobile industry and making them smarter is more crucial than ever. GM is showing the world how that's done.

Robert LeBlanc is senior vice president, Middleware Software, for IBM Software Group.

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish