How things have changed: that little RFID tag that sets off an alarm if it’s not removed by a sales associate has evolved to become the starting point of an intricate Internet of Things data analytics ecosystem that is changing the entire retail business model.
So much so, that according to McKinsey & Company, the economic impact could be as much as $410 billion to $1.2 trillion per year by 2025.1
At their most basic level, passive RFID tags comprise a memory and controller IC, a coil and an antenna at one end, and an RFID reader, or interrogator at the other end (Figure 1.)
Figure 1. A basic RFID system is simple to design and implement, but as integration increases and power consumption decreases, they can perform increasingly complex and useful functions, including communicating ambient parameters and movement. (Image courtesy of Impinj.)
When the reader sends out a coded interrogation signal, the electromagnetic (EM) waves are picked up by the coil, which converts the signal to electricity sufficient to power the IC and enable it to transmit the stored information over the antenna and back to the reader.
At first, RFID was useful simply to help prevent theft, and it seemed that was to be the extent of their use. Later, asset tracking became a critical application as players in the supply chain learned it was easier to track a non-line-of-sight RF signal instead of barcodes. A key enabler of this evolution was the advancing integration of electronic components and their falling costs.
This allowed the deployment of multiple RFID readers in stores without adding significantly to overhead, which allowed asset – as well as staff -- movement to be accurately tracked within a specified range or enclosed environment.
All this is good, but it’s when RFID and asset tracking are combined with IoT concepts, that things get really interesting. Adding a temperature or humidity sensor to the RFID tag allows the environment to be monitored so perishable goods can be better preserved. Integrating a microelectromechanical systems (MEMS) sensor into an RFID tag allows movement to be detected more accurately and quickly. If a garment or expensive store item is being sampled or dropped, customer service can quickly respond to help customers find what they need – or address whatever damage may have occurred.
The IoT angle to all this is that store and franchise owners can gather numerous data points on assets and inventory and analyze this data in real time to respond to inventory and immediate customer needs, as well as over the long term to optimize store layout and product placement.
According to McKinsey, optimized inventory can lower inventory carrying costs by up to 10%, while optimized inventory can increase productivity by 5%. But what about inventory “shrinkage,” that euphemism for theft and “misplacement” of items? This amounts to a $42 billion problem for retailers in the U.S. alone, according to Chain Storage Age. With the deterrent effect of customers and staff knowing the asset is being tracked, combined with rapid awareness of suspicious asset movement, McKinsey believes IoT could reduce losses by one half to one percent of the cost of goods sold. This translates to anywhere from $23 to $92 billion saved per year in 2025.
The good thing is that much of the equipment and infrastructure to enable store owners to start implementing RFID-based asset tracking and IoT is already available and is being put in place (Figure 2.)
Figure 2. The Intel Retail Sensor Platform is starting with RFID but adding Wi-Fi and Bluetooth communications as well as video capability, all of which can be securely aggregated via a local gateway and the data analyzed in the cloud.
For example, Intel has its Retail Sensor Platform for retailers and system integrators. The Platform is an “out-of-the-box” solution that is easy to install and manages RFID and other sensors through a gateway, while supporting analytics capabilities from the edge to the cloud. Big data analytics are supported with the Trusted Analytics Platform (TAP).
Note that I mentioned “other sensors” above: Intel promises to support other communications interfaces and protocols beyond RFID, including video, Wi-Fi, and Bluetooth low energy. All that data can then aggregated and analyzed.
The software is industry standard and open, and users, developers, retailers and system integrators can write applications to the external API and analytics engine to get access remotely to the gateway.
Implications for bricks and mortar
While the analytics will go a long way toward helping retailers improve both their top- and bottom-line performance, it may also be a contributing factor to the up-ending the retail model with stores turning into “hyper-local” viewing and distribution facilities. These may further lower staffing costs as they’ll have minimal requirement for sales associates, as much of the purchasing may have already occurred online. The facility may only be needed for pickup.
How this evolves is yet to be determined: it may swing completely one way toward completely automated retailing facilities, or it may find a middle ground with a mix of retail brick-and-mortar, hyper-local distribution houses with robotic attendees, and online shopping with shipping to the home.
We’ll have to do the analytics to find out.
1: Increasing Profitability for Brick and Mortar Retailers