loT (Internet of Things) is interconnected computing devices forming a system. These devices swap data networks sans human interaction. Multiple IoT applications are under development like smart city, air pollution check, and e-health. Technology is ubiquitous today.
Supermarkets are a defining contemporary feature. They sell numerous commodities and we can take our pick. However, it is observed that customers waste substantial time locating the products they want to buy. The smart supermarket eliminates such difficulties and customers can effortlessly shop and checkout without any hassle.
Fig: Smart Retail
RFID technology lies at the core of IoT. It links physical asset data to process flow data. The RFID-based process nodes are associated with all physical asset data. For example, RFID-equipped machine tools monitor all physical asset data. When it comes to the operating mechanisms, the RFID tags are attached to their products.
Sensors: These can be thought of as a system's ears and eyes as they ascertain the immediate environment, event(s) occurrence(s), and send the collated information. Sensors monitor and perceive physical phenomena or events. Every sensor is categorized on three specifications: sensor type, sensor methodology, and parameters. Sensor type describes the sensor genre, i.e., whether the product is a heterogeneous or homogeneous sensor or a multidimensional or single-dimensional sensor. The methodology comprises the ways (active or passive) through which a specific sensor gathers information. Sensing parameters are the total number of variables that a sensor can sense.
Gateway: All information sensed and collated by sensors is sent to the gateways. These function as common contact points where different information pouring in from heterogeneous sensors gets together. A gateway oversees the global address of an individual store through the use of IPv4 addresses. The coverage area allotted to each gateway in each store with a distinctive IP address enables products sold at that particular store to be easily identified. Every gateway is assigned multiple stores to amplify the granularity of store discernment. A gateway follows its adjacent gateways and also the total population of stores working in its area. Last but not the least, the gateway sends all kinds of unstructured information to successive cloud servers.
Cloud: It works as a storage unit and processing component for the full system. Customer demands are answered at the cloud part. It tracks all stores located within a shopping complex and details all products being sold at their respective stores. It conserves a database where information related to individual products like product ID, date of manufacture, product name, and RFID tag serial number is stored. The scanned serial numbers, during product purchase, are transmitted to the cloud, complete with their payment details. RFID readers are duly informed after completion of a purchase and the product status of those respective products is updated.
Mobile Device: It functions as a customer representative. A mobile device permits the customer to search products and purchase them on a real-time basis. Any special offer or notification appropriate to a specific product is communicated to a mobile device and an appropriate response is awaited.
If smartness involves collating data and communicating with suppliers and customers, then the store shelves and their potential for intelligence come center-stage. A smart shelf will revolutionize the service given by retailers not only to shoppers but also manufacturers as well. Shelf interaction with a shopper will be natural, respectful, and comfortable on matters of shopper privacy. A smart shelf senses emotional states and comprehends natural human language and context. Shoppers will be intelligently served after being assessed whether they are relaxed, in the mood for discovery, stressed, in a hurry, confused, or near making their purchase decisions.
As per the Intel report, shelving can be rationalized into three categories: 'Good category,' 'Better category,' and 'Best category. ' Good shelves come with simple sensing and base display capabilities and zero communications. Better shelves come with more sophisticated sensors, local intelligence, and improved cloud interaction. The Proximity sensors will be substituted by multiple capabilities like seeing, smelling, feeling, understanding, and intuition of the ecosystem around them. These sensors will ascertain the contents they have using cameras, weight sensors, RFID readers, or other technologies. They will simultaneously use 3D cameras, touch and proximity detectors, and microphones to interact with the shopper. Together with local computing resources, these devices will eliminate any cloud processing requirements with their privacy concerns.
The best smart shelves will add increased cloud-based resources to such capabilities to supply product information, personalized shopping guidance, social media reviews, and discounts for the individual shopper. Promotional offers, along with personalized or dynamic pricing, are also possible. A few shelves may harness the shopper's smartphone – the processor, display, touch screen, and microphone – as part of total interaction. Others may use their hardware. These may include OLED/LED high-quality video displays or even holographic devices. Display deployment, however, will be carefully considered, as an excess of bright screens in a small area may turn out to be overwhelming.
Then again, fun and easy-to-use interfaces – inclusive of touch, gesture, virtual, or augmented reality – could pave the way for two-way conversations and highly attractive customization actions.
Smart shelves with adequate computing power can engage in fruitful conversation and offer appropriate responses. A good example is responding to a shopper's question about where to ferret out a product. The same shelf can also make further compatible suggestions to the customer, based on its acquired data related to its purchasing history.
Smart retailing manufacturers will profit from insights into linger times, customer demographics, traffic, and advertising success. Machine vision-equipped cameras and shelf sensors may constantly track inventory levels and schedule replenishments with a dip in product quantity. These sensors can also track shoplifting attempts. Customer behavior gathered by the combo of store movement trackers and shelf sensors are dispatched to a cloud analytics resource and subsequently integrated with collated data linked to online behavior – comprising clicks, wish lists, hovers, and shopping carts.
Low-cost RFID tag alternatives like passive UHF temperature-sensing RFID inlays preserve perishable foods. The on-chip temperature-sensing circuit of such SMARTRAC devices can digitize the temperature reading of product temperatures into 12-bit numbers. A UHF reader can scrutinize, accompanied by the tag's distinctive identifier. Simple moisture-sensing capabilities are present, dependent on the measurement of impedance variations.
Retail operators are keen to source integrated solutions containing components that collaborate to offer an improved shopper experience while collecting and analyzing relevant retail data for better management.
Each shelf comprising this system comes armed with a meshed sensor array that includes photodetectors, strain sensors, spillage detectors, and microphones, along with a data processing system for managing sensor signals. This sensor mesh layer is latched to the bottom of every individual shelf. Such a system additionally includes a video display set for showing the attributes of those specific products being retailed from the particular shelf. An arrangement of wireless transmitters delivers these attributes.
Storage devices come in many flavors, with user input devices comprising a judicious mix of a keypad, keyboard, mouse, and devices for image capture, motion sensing, smell detection, light detection, microphone, or fused devices containing multiple functions.
The central server can offer access information like a pager, email addresses, or SMS for employees. This information enables them to be contacted and consequently updated about undesirable situations like low stock.
The video displays may exhibit product characteristics like price, chemical freshness, and weight regulated by methane or color emission, expiration dates, nutritional values, calories, recipes, and other information as needed. They also display promotions of compatible possible interest products.
Microphones can track sounds that divulge the expansion of a container when its contents are compromised or handled inappropriately on the shelves. Such microphones are fitted using piezoelectric material strips that produce a signal on the detection of minute vibrations. Various sensors are integrated into a single sheet incorporating one layer reserved for a strain detector, the next layer for a photodetector, another layer for an acoustic sensor, and so on as needed.