Modern　smartphones　equipped　with　various　sensors　provide　a　cost-effective　solution　for　urban　pavement　ride　quality　monitoring.　In　this　field,　road-induced　vehicle　vibrations　collected　by　smartphone　accelerometer　are　quantified　following　ISO　2631-1　as　occupant-perceived　driving　comfort　intensity,　serving　as　a　guideline　for　evaluating　ride　quality.　However,　existing　methods　often　ignore　the　heterogeneous　effects　of　vehicle　suspensions　and　speeds　on　driving　comfort　intensity,　challenging　the　objectivity　of　ride　quality　monitoring　results.　Therefore,　this　study　advocates　a　multi-vehicle　collaborative　monitoring　scheme　and　proposes　a　tailored　two-stage　method.　In　the　first　stage,　multi-vehicle　vibrations　along　the　same　path　are　collected　via　smartphone　and　transformed　into　multi-source　driving　comfort　intensity　(MSDCI)　information.　The　second　stage　introduces　a　probability　fusion　method　based　on　Generalised　Extreme　Value　(GEV)　theory　to　model　MSDCI　information　as　an　uncertain　event,　further　categorised　into　six　levels　of　driving　uncomfortable　events　to　jointly　characterise　ride　quality　and　develop　a　warning　metric.　The　method　was　validated　through　a　real-world　case,　showing　that:　(1)　the　GEV-based　method　is　robust　in　fusing　MSDCI　information;　(2)　pavement　locations　with　severe　ride　quality　problems　can　be　identified　by　the　warning　metric.　The　proposed　method　contributes　to　sustainable　urban　pavement　management　and　will　be　promoted　by　integrating　crowdsourcing　technology　.