Determination of the levels of heavy metal ions would support assessment of sources and pathways of water pollution. However, traditional spatial assessment by manual sampling and off-site detection in the laboratory is expensive and time-consuming and requires trained personnel. Aiming to fill the gap between on-site automatic approaches and laboratory techniques, we developed an autonomous sensing boat for on site heavy metal detection using square-wave anodic stripping voltammetry. A fluidic sensing system was developed to integrate into the boat as the critical sensing component and could detect <= 1 mu g/L Pb, <= 6 mu g/L Cu, and <= 71 mu g/L Cd simultaneously in the laboratory. Once its integration was completed, the autonomous sensing boat was tested in the field, demonstrating its ability to distinguish the highest concentration of Pb in an effluent of a galena-enriched mine compared to those at other sites in the stream (Osor Stream, Girona, Spain).