Competition in the automotive industry is intense, and successful companies must constantly innovate by introducing new technology to differentiate and improve their brands. As a result, today’s vehicles have evolved from mechanical devices into complex integrated technology platforms with embedded software powering all major systems including: engine control, power train, braking and airbags, driver assistance, and infotainment.
Consider these incredible statistics: today’s vehicles have more processing power than any of NASA’s early spacecraft, including the Apollo lunar lander. A modern vehicle has more lines of code (10M) than the latest technology in the US Joint Strike Fighter.
These technological changes are affecting automotive brands in subtle ways, as consumer opinions are influenced by their interaction with this technology. In many ways the brand is the user experience, more so than the technical specifications.
Technology Defines the Brand
A strong brand creates significant value in all industries, and that is certainly the case with automobiles. According to Interbrand’s 15th annual Best Global Brands Report, a focus on integrated technology is helping to drive brand loyalty and value for automakers. In the most recent survey conducted, the collective value of the automotive brands appearing on the Global Brands ranking increased 14.6 percent. Three out of the five “Top Risers” listed were from the automotive sector, making the 15th annual report a record-breaking one for the auto industry, according to Interbrands Best Global Brands Report.
As automobiles evolve from mechanical to software-enabled devices, automakers must rethink fundamental product development principles, including moving from a sequential design process to a more agile approach, with higher degrees of collaboration between self-directed, cross-functional teams.
Software Quality Must Be a Priority
Now, more than ever, software quality needs to be at the top of the list for major auto brands looking to preserve – and elevate – brand status.
Statistics show that more than 50 percent of auto recalls are now due to software bugs, not mechanical issues,, according to GeekWire. With an industry average of 1 bug per thousand lines of code, the average car has 10,000 latent bugs.
What is the Goal of Testing?
The first step to improved quality is rethinking the role of testing in the development cycle. Often, the release cycle is seen as a zero sum game, where investing more time in testing simply pushes out release dates by the same amount. This mindset neglects the value of these test artifacts over the lifecycle of the product.
This flawed thinking is deeply ingrained in the industry. The first step for organizations searching for improved quality is to start thinking in terms of total lifecycle cost, not simply time to release 1.0. Most organizations have a huge quantity of legacy tests, but often they are not being leveraged properly. The following are common flaws in the implementation of testing:
- Testing completeness is not measured
- Tests are not easily shared between team members
- Testing is not automated and continuous
Measuring Testing Completeness
Source code coverage analysis is an easy-to-implement way to measure the completeness of existing tests. Measuring code coverage for all types of testing and combining this into a single view provides valuable insight into under-tested and over-tested areas of the application. By understanding gaps in existing tests, users are able to improve testing efficiency, and over time, fill these gaps with new tests. The result is improved application quality.
Sharing Tests across Teams
Tests are often compartmentalized, with the majority of tests being owned by the Software Quality Assurance (SQA) team. If developers do not have access to these tests, then it is not surprising that bugs are found by SQA late in the release cycle — after new features are integrated and these tests get run for the first time. The lack of test sharing can also result in an adversarial relationship between developers and SQA as much time is wasted by developers trying to understand bugs, and SQA trying to diagnose test failures. Sharing tests offers a variety of benefits to the whole team:
- Bugs will be fixed on the day they are introduced, not weeks later
- Code changes that break existing tests will not be integrated
- Team members will focus on edge cases versus nominal path testing
- SQA will have more time to write new tests
Automated and Continuous Testing
The key to creating a sharable testing infrastructure is to ensure that tests are small and run fast, execution is automated and dependable, and results are easy to understand.
Automated and continuous testing will only work if a common testing platform is implemented across the organization that allows all team members to run existing tests and integrate new tests easily. The tests must capture all preconditions, test values, and expected results so that test execution is as simple as clicking a button and viewing a pass or fail status; and most importantly it must be easy for developers to debug failing tests.
Conclusion
In the 1950’s, a revolution in manufacturing quality was initiated by W. Edwards Deming, who taught companies how to produce higher quality products at the lowest cost possible. Central to Deming’s philosophy was his criticism of the dominant form of quality control in use at the time: products were inspected for defects only after they were made. In contrast, Deming maintained it was better to design the manufacturing process to ensure that quality products were created from the start.
This seems obvious to us now when we think of manufacturing, but the software industry has been slow to adopt a similar philosophy.
Over the last 50 years, automotive suppliers have revamped design and production processes to control cost, drive quality, and create brand loyalty. These gains may be lost if the same focus on quality is not applied the software that controls the majority of systems in modern automobiles.