In this paper we provide an intelligent and convenient shopping cart, Intelligent Shopping Assistant System (ISAS). With ISAS, customers can concentrate on what he/she has to buy, and do not be bothered where he/she has to buy or it is want to buy. In contrast to conventional shopping cart, two modes of autonomous functions are added to ISAS in order to reduce the labor of customers for pushing goods in the cart. First, ISAS can follow customers by steadily recognition and tracking, then customers could get more freedom which they didn't need push traditional shopping cart by themselves. Second, ISAS can go to products that customers want to buy by localization and navigation. In addition, Q-learning obstacles avoidance will execute when ISAS around is full of obstacles or people. For convenience of customers, they could construct shopping list by ISAS app. with just a few action. Also with two technic combination of data mining and deep learning, ISAS will provide products information and let customer decide it is want to buy.