Project: Real-time Information Application

_x000D_The goal of RETINA project is to develop a marketing tool that measures success of marketing activities in retail stores through the analysis of both online and offline information sources and helps management build marketing strategies that best fit to their customers' attributes and shopping patterns._x000D__x000D_Offline information is collected at the stores and comprises biometric data from customers and customer behavior data in the stores._x000D_Biometric data is acquired through face recognition techniques applied over video surveillance cameras pointed to key areas of the store such as entrance, shop windows, cash desk and others (e.g. access to the changing rooms area in clothing stores). Biometric data include attributes such as gender, height, age group and constitution. These attributes are used to classify customers into different predefined profiles._x000D_Customer behavior data is acquired through people counting systems that identify number of visitors to the store and their paths inside it, measuring time that visitors spend in each area of the store and identifying hotspots. All this information is obtained through zenithal cameras that cover the store plan._x000D__x000D_Online information is collected from the customer website, customer communications & fansite, social networks such as Facebook, Twitter and others, and other online resources (e.g. Google Analytics). Data gathering process includes a semantic analysis of comments published by customers/followers in social networks and customer sites in the Internet (promotional website and fansite) as well as the analysis of the amount of followers, comments and clicks on product pages on customer internet resources._x000D_All the information gathered from both offline and online sources is aggregated and analyzed to provide management with real-time information on the effectiveness of marketing activities put in place and their adaptation to the customer profiles that are visiting the store._x000D__x000D_RETINA gives management real-time information of morphologic characteristics from people who stay in front of the shop windows and the amount of time they stay in front of them. It also measures how many people passing by the store entrance decide to enter the store and which are their morphological attributes. Once inside the store, RETINA tracks every visitor (general morphological features, no sensible info) and registers visitor paths, identifying hotspots. Key areas of the shop are covered by face recognition cameras that again provide morphological attributes of visitors standing in front of key displays._x000D_The face recognition process is initiated at the cash desk when the sale process begins, linking morphological attributes to ticket information and thus linking customer profile with store, shop assistant, date and time, purchased items, payment method, vouchers / offers applied and fidelity program information (if applicable)._x000D_Company's information systems would also become data source._x000D__x000D_Clothing stores get the most from RETINA for example if clothes were labeled with RFID tags , RFID sensors placed at the entrance of the changing room area would detect a customer taking items to the changing room, the face recognition process would initiate to link customer morphological attributes to items taken to changing rooms. _x000D__x000D_The core system evaluates offline and online information and delivers the following results on a real-time schema:_x000D_- Degree of success of marketing activities and campaigns put in place by the customer. This functionality is obtained from the development of complex algorithms that evaluate changes in shopping patterns from customers and followers response in customer's online resources._x000D_- Characterization of customers, store visitors and online followers by their morphological attributes and other attributes obtained from their personal profiles in the online resources._x000D_- Multidimensional analysis of sales and their association to customer profiles. It allows measuring campaign success as a whole (sales increase) but also tracking campaign success on specific targets._x000D__x000D_Due to the fact that customers' needs vary from one to another based on the retail subsector and customer marketing strategy, RETINA is conceived as a modular system. According to this, customers can find the module combination that best fits their needs._x000D__x000D_The consortium is composed by two Ps as described below:_x000D_- CSD (coordinator). Spanish R&D intensive SME. It will be in charge of the development of the core application, customer data analysis and development of complex algorithms to measure campaign success. CSD will subcontract an industrial optical R&D Institute to develop face recognition functionalities (futher info in point 2.4.1)._x000D_- InfoDif (P). Turkish R&D intensive SME that will provide the consortium with vast know-how on image processing and will perform people counting and path identification processes as well as RFID related issues._x000D_

Acronym RETINA (Reference Number: 6624)
Duration 01/11/2011 - 30/06/2014
Project Topic RETINA is a marketing information system that provides management with real-time behavioral information of customers and shop visitors. Among information sources will be the analysis of images of CCTV systems placed on shops and the results of semantic evaluation of comments on social networks.
Network Eurostars
Call Eurostars Cut-Off 6

Project partner

Number Name Role Country
2 InfoDif Yazilim ve Bilisim Teknolojileri Co. Ltd. Partner Turkey
2 Company for Software and Development S.A. Coordinator Spain