客戶購物體驗提升之智能購物車設計

Abstract

商店內的顧客行為追蹤是一個新興的領域，在此領域，已有許多的相關研究。現有追蹤正在購物的顧客行為的方法，已經使用了攝影機、智慧型手機及智慧型購物車 (Smart Shopping Cart, SSC)。目前這些方法皆有其侷限之處。為了克服這些限制，我們提出了一個獨特的SSC設計方法。在此方法中，我們設計並g實作基於Raspberry Pi的智能購物車雛形，其特性具有SSC追蹤、排隊辨識及隊伍特性估算。藍芽信標(BLE Beacon)技術和項目記錄用於SSC追蹤，可以透過取得SSC的微共振腔特徵及BLE Beacon 資訊時辨識隊伍。隊伍特性估算包括服務時間及等待時間的估算，對於隊伍特性估算，我們考慮了每個SSC中的項目數量及隊伍特性的歷史資訊。我們在SSC中整合了許多項功能來加以改善顧客購物過程，我們並且考慮多個實際情況進而做了大量的實驗來驗證。

In-store customer b ehavior tracking is a new area. On this topic, a large amount of work has b een done in this direction. To track the customer activities on shopping existing approaches have used cameras, smart phones, and smart shopping carts (abrv. SSC). All these approaches have their own limitations. To overcome these limitations we came up with our unique SSC design. In this work, we design and implement a prototyp e of smart shopping carts based on Raspb erry Pi, which features SSC tracking, Queue recognition, and Queue prop erty estimation. BLE b eacon technology and item logs are used for SSC tracking. Queue recognition is done with the help of SSC micro-cavity features and BLE b eacon proximity information. Queue prop erty estimation includes estimation of service time and waiting time estimation. For queue prop erty estimation we take into consideration, the numb er of items in every SSC and historical information of queue prop erties. We integrate several functionalities in SSC to improve customer shopping exp erience. We have also done an extensive amount of exp eriments by considering several practical scenarios.