Edge computing and data optimization for remote diagnostics of vehicles
Vitalii Bonar, Volodymyr Hotovych, Liubomyr P. Matiichuk · Scientific journal of the Ternopil national technical university · 2025
This paper proposes the use of edge computing for remote vehicle diagnostics. Instead of continuously streaming raw telemetry, the on-board computer performs simplified analytics, adaptive sampling, and data compression, so that the outgoing network channel transmits event packets and short periodic summaries (summaries) of data. The system design focuses on three goals: to maintain the ability to make important diagnostic decisions, to keep the reconstruction error acceptable for analysis, and to notify of a fault even with weak or variable connectivity in a reasonable time. The paper proposes a simple and reproducible way to account for data volume and latency, shows how to configure basic controls for each signal, and discusses the use of multi-peripheral edge computing (MEC) to reduce latency. The result is significant savings in outbound traffic without compromising diagnostic capability.