A study of Bluetooth-derived journey times from Greater Manchester, published in European Transport Research Review and carried out by Mohammadmahdi Rahimiasl, Ynte Vanderhoydonc and Siegfried Mercelis within the European Union’s Horizon 2020 TANGENT project, shows that classical denoising filters designed for entire datasets degrade predictions when applied to the few hours of recent observations available to live forecasting systems. The researchers instead train a deep learning model on raw, noisy windows to predict what an expert-selected filter would have produced, folding denoising and forecasting into one causal forward pass that meets real-time latency requirements. The data’s properties are challenging: Bluetooth sensors detect device MAC addresses without knowing the transport mode, match counts fluctuate over time, and physical obstructions can shrink detection range, producing journey-time series with missing values, outliers and sharp fluctuations. Across the 204 predefined routes recorded at 15-minute intervals over 2022, roughly 20 percent of entries are missing, and aggregated statistics show a positively skewed, heavy-tailed distribution with substantial outlier influence. Citing an external report, the authors note that drivers in London lost an average of 101 hours to congestion in 2024, a 2 percent increase from the previous year, while across the UK the average was 62 hours, with an estimated national cost of 7.7 billion pounds in lost time. The paper identifies that wavelet shrinkage denoising and Butterworth low-pass filtering were conceived for full