Abstract: AIS（Automatic Identification System） data is used to improve the location selection of offshore buoy platforms through introduction of a GIS（Geographic Information System）-based kernel density estimation algorithm. A kernel density model of marine traffic flow is built to calculate the traffic density in candidate water areas and find the characteristics of the traffic flow in each area. The traffic flow is graded on density basis and displayed in different colors. The traffic density map of every candidate location is processed on monthly basis and the suitable buoy locations are picked up. Final location selection suggestion is made after examining the navigational environment of these locations. The effectiveness of the algorithm is demonstrated through a practical engineering case.