Data Fusion and Sensor Integration Across Automotive and Rail Systems

The ambitious scope of data fusion across vehicle categories is driven by common needs for resilient physical sensing capabilities in an increasingly complex cyber-physical system context. The construction of advanced maps, improved safety and integrity of operations, enhanced decision-making and autonomy, advanced shared-vehicle concepts, asset health monitoring, and predictive maintenance of critical assets all require sensor integration across multiple sensing modalities, operator domains, and levels of operator control. Data fusion is indeed an overarching trend across automotive and rail systems. The fundamentals of data fusion theory, terminology, levels, and architectural aspects apply to both domains, and many of the new developments found in one domain can be relevant for the other. Nevertheless, the substantial and specific application needs only found for automotive or rail systems cannot be neglected. Such needs provide the essential motivation for concentrating research efforts into the two domains when applying specific data fusion techniques or algorithms.