SUMMARY REMARKS & TESTBEDS
The ThinkINFIRE+ experiment aims at developing and experimenting on FIRE+ testing facilities a new indoor analytics service (ThinkIN+) for retail businesses. The service will allow customers to track in a privacy-aware fashion how shoppers move inside their stores and interact with items on displays. The service will make use of Internet of Things devices to collect relevant data on in-store shopper behaviour, which will then be processed in the cloud to extract relevant KPIs and actionable metrics. The proposer, ThinkINside srl, is already active in the field of in-store analytics, and is experiencing requests by its customer base to provide an enhanced service able to provide spatial analytics at the level of single store areas. This new service would require additional data-intensive sensors to be installed, and a new pipeline for the real-time processing of data. A proof-of-concept implementation is already available, but it requires being re-engineered leveraging state-of-the-art big data frameworks and experimented and tested in a realistic large-scale setting, in particular for assessing its scalability. The objectives of the ThinkINFire+ project are to re-engineer the service prototype using state-of-art tools, to deploy it on FIRE+ and to experimentally assess its performance by running large-scale tests. The Tengu platform will be used as both a playground (guiding the re-engineering thanks to its pre-packaged big data tools) as well as test environment (to assess scalability in terms of number of stores that can be supported by a single cloud instance). The final outcome will be a validated, market-ready implementation that will be commercialised by the applicant.In carrying out the experiment we will extensively test the functionality and features of the FIRE+ facilities, providing feedback to FIRE+ stakeholders and facility owners on the benefits and suitability of FIRE+ to act as one-stop-shop for the development and experimental validation of IoT/Big Data platforms.