Kimio Nishino

PLE for automotive braking system with management of impacts from equipment interactions

We report here an industrial application of the Product Line Engineering (PLE) for the development of electronic braking systems. The cost of software engineering in automotive control systems is increasing as new functions for safety, comfort, and improved fuel efficiency are integrated into electronic control units. Therefore, Component suppliers for automotive control systems adapt their products to the requirements of car manufacturers by modifying the software specifications, such that it makes minimal changes to the mechanical structure and the electrical and electronic (E/E) components hence reduces the cost. PLE is an effective approach to manage or even reduce the software variations resulting from these modifications. However, one problem is that the software specifications of automotive control systems need to be redesigned after system testing with vehicles. This is because vehicles consist of many mechanical parts manufactured by different suppliers, and the characteristics of the parts can interact with each other. This problem makes it difficult to reap the full benefits of PLE. We propose an approach to analyze the potential impact from such interactions by using a system model that expresses the system architecture that includes the parts of different suppliers. Based on this model, the software architecture was designed to localize the impact to several software components. Additionally, a feature model was designed to the enable management of the localized impact by expressing it as variability. This method helps software engineers specify the software components that can have an effect on the actual equipment, and determine which modifications to the software specifications are necessary. We applied PLE with the proposed method in the development of electronic brake control system. We confirmed that our approach greatly increased the efficiency of PLE for the development of such automotive control systems.