Silent emissions: The cyber-infrastructure environmental impacts of autonomous vehicles

In the consideration of how autonomous vehicles (AVs) can contribute to or hinder in achieving environmental goals, there is a critical aspect that must be included: AV data management. This aspect, encompassing both in-vehicle and cyber infrastructure energy requirements, has been neglected in many AV environmental impact studies. To address this gap, we calculated the energy requirements and greenhouse gas emissions associated with AV data management, including previously neglected cyber infrastructure components such as data storage and security. With a mathematical model, we tested various scenarios within a United States-based AV fleet, consisting of 15 key variables like computational efficiency, vehicle models (e.g., Waymo, Tesla), sensor capabilities, data transfer, storage, and cybersecurity. Our findings reveal that neglecting to include cyber infrastructure components underestimates the environmental impact of AV data management by at least 20 %. In the context of increased adoption, a 40 % market penetration of fully autonomous vehicles in the United States could demand up to 6278 GWh/year for data management alone (about 4.3 % of current US solar capacity for reference), highlighting the potential for AV adoption to intensify energy needs amidst the renewable energy transition. Thus, our results emphasize the significance of including the essential cyber infrastructure for AV data management in environmental assessments.