The modern vehicle is increasingly described as a "computer on wheels," equipped with high-performance processors and massive storage to handle the demands of autonomous driving. Yet, for the vast majority of their lifespan, these sophisticated machines sit idle in parking lots or garages. Researchers are now proposing to harvest this latent capacity, transforming stationary cars into a decentralized, opportunistic edge computing layer.
The Mobility Dilemma at the Edge
Traditional Mobile Edge Computing (MEC) relies on fixed Roadside Units (RSUs) to process data for nearby vehicles. However, as Jun Zhang and Khaled B. Letaief argue in Mobile Edge Intelligence and Computing for the Internet of Vehicles, the "data explosion" from autonomous sensors is quickly outstripping the capacity of fixed infrastructure.
While moving vehicles can theoretically share resources via Vehicle-to-Vehicle (V2V) links, their high mobility creates a "mobility dilemma." As Muhammad Waqas and Zhu Han explain in Mobility-Aware Fog Computing in Dynamic Environments, the fleeting nature of encounters between moving cars makes task scheduling unreliable. If a vehicle moves out of range before a computation is finished, the task fails.
Parked Vehicles as Stable Fog Nodes
The solution lies in the high density of stationary vehicles in urban environments. By repurposing parked vehicles as "temporary fog nodes," the network gains a resource pool that is both high-capacity and geographically stable. Dusit Niyato and colleagues in Computation Offloading and Content Caching Delivery in Vehicular Edge Computing highlight that parked cars offer a more reliable alternative to moving nodes, effectively bridging the gap between fixed base stations and highly mobile traffic.
This paradigm, often called "Fog Computing on Wheels" by researchers like Tom H. Luan and Peng Cheng, treats the city's parking infrastructure as a distributed data center. In Fogbanks: Future Dynamic Vehicular Fog Banks for Processing, Sensing and Storage in 6G, Amal A. Alahmadi and Jaafar M. H. Elmirghani demonstrate that orchestrating these "mobile micro data centers" can reduce total network power consumption by up to 80% compared to centralized cloud architectures.
Harvesting Idle Resources
The intuition behind this research direction is based on three key observations:
- Underutilization: Vehicles are stationary for roughly 95% of their lives. During this time, their onboard computers—capable of trillions of operations per second—are entirely idle.
- Energy Availability: Electric vehicles (EVs), in particular, possess large battery reserves. Jiming Chen and Haoyu Li note in Smart and efficient EV charging navigation scheme in vehicular edge computing networks that this power can be leveraged during charging periods to offset travel costs for the owner.
- Proximity: In dense urban areas, a moving vehicle is almost always within range of a parking lot or a line of street-parked cars. Guoling Liang and Liehuang Zhu propose a collaborative framework in Task Offloading for Vehicular Edge Computing Based on Improved Hotstuff under Parking Assistance that integrates fixed RSUs with these idle parked resources to prevent system bottlenecks.
Toward Vehicle as a Service (VaaS)
This shift represents a move toward the Vehicle as a Service (VaaS) model. As described by Yiqin Deng and Pan Li in Vehicle as a Service (VaaS): Leverage Vehicles to Build Service Networks and Capabilities for Smart Cities, vehicles are no longer just transport machines but ubiquitous infrastructure components.
By leveraging the sensing, computing, and storage power of parked cars, cities can build a "SCCSI" (Sensing, Communication, Computing, Storage, and Intelligence) network. This decentralized approach, also echoed in the ADROIT6G project by Christophoros Christophorou, allows the network to dynamically expand its coverage and computational power by harvesting unutilized local resources at the "far edge."
Ultimately, the goal is to turn traffic congestion and parking into a localized computing resource. As Ahmad Alhilal and Pan Hui suggest in Distributed Vehicular Computing at the Dawn of 5G, this mechanism transforms what was once a source of urban frustration into a vital component of the smart city's digital backbone.