The work aims to design a self-adaption electric packaging machine to increase the working efficiency of e-commerce, express and logistics and collect the data related to packages, thus providing data for intelligent logistics at the industrial front line. The problems of packaging machine currently sold in the market and the problems to be solved in express logistics industry were analyzed. In combination with the difficulties and the principle of packaging, the mechanical structure and control scheme of the packaging machine were designed, and the working efficiency was calculated theoretically. Then, calculation results were verified by experiments. The prototype could be used to seal cartons of any length (110~320 mm wide and 153~290 mm high) and required time was 5~12 s. At the same time, the machine could collect the detailed data of the package and upload these data to the host computer. The design can automatically adapt to pack carton in different sizes. Features like pure electric, small size and low cost are suitable for small and medium-sized e-commerce and express delivery sites. The technology can also be applied to large B2C e-commerce, such as JD, Tmall and other terminal goods packaging and sealing links. The relevant logistics data collected will be the big data foundation for building an intelligent logistics system in the future.