The four major mobility trends: Electrification, Automation, Connectivity and Ride Sharing drive technological and societal implications, resulting in the emergence of a new ecosystem that could offer faster, cheaper, cleaner, safer, more efficient transportation. Connectivity and Automation capitalize on the merits of the Electrified vehicle platforms.
The potential safety related benefits of Connected Automated Vehicles (CAV) are relatively well quantified and accepted, but there is high uncertainty related to their Energy Consumption and Emissions implications, justifying continuous in-depth analysis. SAE International defined five levels of automation, requiring the vehicle to perform specific Advanced Driver Assist System (ADAS) functionality based on the inputs from a complex sensor suite. The CAV energy and efficiency analysis methodology follows the same general approach as for conventional vehicles but includes several specific aspects.
In a CAV, the sensor suite and compute stack power consumption is an extremely important component of the CAV-Centric intrinsic energy efficiency (HW and SW dependent). The CAV’s perception sensor suite and compute stack perform complex, power intense tasks. Lidar is the primary sensor for the advanced perception functions, influencing the CAV power and energy requirements.
This paper demonstrates that by optimizing the sensor and compute stack power consumption at the system level, a complete sensor suite covering all CAV perception needs can require less wattage than an individual sensor. Model based simulation and control strategies can be applied to all CAV Systems (example engine, transmission, driveline), over various drive cycles and considering the infrastructure. Physics based, deterministic simulation approaches improve calibration and control speed, accuracy, and capture better the dynamics of all CAV systems in transient scenarios.
The standardization of the CAV components and systems offers multiple benefits, including the proper characterization and predictability required for efficient model-based controls. The simulation and control of the electrified CAV require continuous efforts for developing the plant models of the powertrain and perception stack. This paper brings a contribution to these aspects and emphasizes also the long-term benefits resulting from the standardization of validation protocols for the discrete components and overall CAV architecture.