Forskningsradar
← Tech & AI
Tech & AI 3.1

Different carmakers can now safely share road friction data

Researchers have validated that tire-road friction estimates from competing vehicle suppliers can be reliably compared and combined, opening the door to cross-brand data sharing. The finding, tested across 100,000 kilometers in Swedish road conditions, removes a major technical barrier to unified safety systems that could reduce traffic accidents.

Originaltitel: Fusion of connected vehicle data from multiple suppliers estimating tire-road friction

Abstrakt

<p>In the digital era, all sectors of society are moving towards increased connectivity, including the transport system. Digitization is essential for achieving Vision Zero, a future without road traffic casualties and with minimal environmental impact. Connected vehicle (CV) data, currently shared within fleets of the same brand, will soon be exchanged between different brands. However, interpreting parameters such as tire-road friction (TRF), estimated using supplier-specific algorithms, remains a challenge. This study investigates similarities and differences between two TRF estimation methods developed by different CV data suppliers. Both approaches use tire slip and internal vehicle signals. Data were collected in Sweden on public roads, covering around 100,000 kilometres, and span two reference periods (April, July, and October in 2023 and 2024) and three winter seasons (December to March) from 2022 to 2025. A comparison using confusion matrices is performed on road segments up to 550 m in length, where 1 h measurements from both suppliers are available. Accuracy in classifying high and low TRF ranges between 85% and 93% during winter. Differences are related to road weather conditions and the suppliers' use of TRF limits. Despite these variations, it is feasible to fuse TRF data from multiple suppliers, particularly for short segments.</p>

Generera ett redaktionellt utkast på svenska