Abstract: In order to obtain traffic data more easily and put the prediction of vehicle speed, a simple-to-use and user-friendly program is developed based on Python, which makes the collection, processing, analysis, prediction and release of road speed data more accessible and integrated. Such goal can be achieved, according to the operation guide of Amap development platform, by taking the following steps: Firstly, through python, a crawler program is designed to collet road speed data. Secondly, these collected data are purified and restored to extract the time series speed data of designated road section. Thirdly, the data are decomposed and used for prediction based on ARIMA model. Fourthly, through Qt Designer, the interface code is devloped and merged with logic code to complete the visual design process including data collection, processing, analysis and prediction. At last, the Django framework is used to complete the Web page of publishing prediction. The program can quickly obtain the speed data of designated road section and provide "congestion forecast" for travelers.