In the retail sector, in recent years, has grown the need to acquire information about customer behavior. This information allows to optimize stores, improve the commercial offer, innovate products and maximize profits by monitoring in real time the behavior and choices of customers. The physical sales point is now no longer the only place where the purchase takes place and for this reason should be able to rapidly align with the needs of customers transforming the spaces and the shopping experience using the data collected. This aim can be achieved through the use of IoT solutions based on machine vision, indoor tracking systems and distributed sensors for environmental parameters monitoring. In this context the research activity had as its main focus the study and development of solutions based on RGB-D cameras, UWB indoor tracking systems and the analysis of distributed networks of capacitive sensors placed inside floor for human movement control. The effectiveness of these measuring systems and the analysis software developed have allowed to realize a complete monitoring system for retail environments that thanks to IoT communication protocols allows to analyze in detail and in real time in a store: how many people, how they move, how interact, what they buy and how long it takes to perform each of the above actions. The research path has been conducted in full collaboration and cooperation with the Grottini Lab company. Through this synergy several installations have been made in real stores that have allowed to test the validity of technological solutions designed and at the same time to collect many data useful for the behavior analys of individual and groups of customers. The data produced with these analysis systems are collected on IoT cloud platforms and once stored, can be processed and made visible in appropriate dashboards in terms of charts and key performance indicator to make a careful analysis of the human behavior in a retail space.
Il settore retail negli ultimi anni ha visto crescere sempre di più la necessità di acquisire informazioni sul comportamento dei clienti e come si muovono all’interno degli store. Queste informazioni consentono di ottimizzare il layout del punto vendita, migliorare l'offerta, innovare prodotti e massimizzare i profitti monitorando in realtime i comportamenti e le scelte dei clienti. Il punto vendita non è più ormai l'unico luogo dove avviene l'acquisto e per questo motivo deve essere in grado di allinearsi velocemente con l'esigenze dei clienti, trasformando gli spazi e l'esperienza di acquisto. Grazie all’utilizzo di soluzioni IoT basate sulla visione artificiale, sistemi di tracking indoor e sensori distribuiti per il monitoraggio ambientale è possibile raggiungere questo obiettivo. L’attività di ricerca ha avuto pertanto come focus principale lo studio ed il test di soluzioni basate sull’utilizzo di camere RGBD, sistemi UWB di tracking indoor e lo studio di reti distribuite di sensori capacitivi posizionati all’interno di pavimenti per il monitoraggio degli spostamenti. La validità dei sistemi di misura utilizzati e degli applicativi software implementati hanno permesso di realizzare un sistema completo di monitoraggio del punto vendita che, grazie all'uso di protocolli IoT di comunicazione, consente di sapere in tempo reale: quante persone entrano, come si muovano, come interagiscono, cosa comprano e quanto tempo impiegano a svolgere ciascuna delle precedenti azioni. Il percorso di ricerca è stato condotto in collaborazione con l’azienda Grottini Lab. Grazie a questa sinergia sono state effettuate numerose installazioni in ambienti di retail reali che hanno consentito di testare la validità delle soluzioni tecnologiche sviluppate ed allo stesso tempo di raccogliere numerosi dati utili all’analisi del comportamento di singoli e gruppi di individui negli ambienti retail.
Machine vision and IoT applications in intelligent retail environments / Contigiani, Marco. - (2017 Mar 23).
Machine vision and IoT applications in intelligent retail environments
CONTIGIANI, MARCO
2017-03-23
Abstract
In the retail sector, in recent years, has grown the need to acquire information about customer behavior. This information allows to optimize stores, improve the commercial offer, innovate products and maximize profits by monitoring in real time the behavior and choices of customers. The physical sales point is now no longer the only place where the purchase takes place and for this reason should be able to rapidly align with the needs of customers transforming the spaces and the shopping experience using the data collected. This aim can be achieved through the use of IoT solutions based on machine vision, indoor tracking systems and distributed sensors for environmental parameters monitoring. In this context the research activity had as its main focus the study and development of solutions based on RGB-D cameras, UWB indoor tracking systems and the analysis of distributed networks of capacitive sensors placed inside floor for human movement control. The effectiveness of these measuring systems and the analysis software developed have allowed to realize a complete monitoring system for retail environments that thanks to IoT communication protocols allows to analyze in detail and in real time in a store: how many people, how they move, how interact, what they buy and how long it takes to perform each of the above actions. The research path has been conducted in full collaboration and cooperation with the Grottini Lab company. Through this synergy several installations have been made in real stores that have allowed to test the validity of technological solutions designed and at the same time to collect many data useful for the behavior analys of individual and groups of customers. The data produced with these analysis systems are collected on IoT cloud platforms and once stored, can be processed and made visible in appropriate dashboards in terms of charts and key performance indicator to make a careful analysis of the human behavior in a retail space.File | Dimensione | Formato | |
---|---|---|---|
tesi_contigiani.pdf
accesso aperto
Descrizione: tesi_contigiani
Tipologia:
Tesi di dottorato
Licenza d'uso:
Creative commons
Dimensione
5.43 MB
Formato
Adobe PDF
|
5.43 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.