The Internet of Things (IoT) allows many kinds of devices (the “things”) to collect and exchange data. While IoT is no longer a new concept, it continues to grow and gain traction. By most estimates, the number of connected devices will have doubled from 2014 to 2017 and will double again by 2020. The number of connected devices varies depending on who you ask, what’s considered a device or how you define “connected,” but the trajectory is clear – more and more of the things around us are talking to each other every day.
This communication, data gathering, and exchange of information between devices can serve many purposes. Devices are now addressable remotely, can interact with users more effectively, are more optimized for performance and functionality, and can cooperate to build increasingly immersive, intuitive experiences for users.
As a data scientist seeking to automate understanding of the physical world by machines, one of the key challenges is in generating data about the physical world in the first place, or “sensing” in the context of the Analytics Stack.
As more and more “things” gain the ability to communicate, one important aspect to consider is the ability of those things to relay information about the environment they are located in. These “things” can become useful sensors, helping to integrate the physical world into our digital systems. The ultimate goal for the input to an analytics platform attempting to make sense of complex real-world activities is to deliver to that platform a complete picture of the world and to let that platform ignore the aspects (or data points) that aren’t relevant, making informed decisions based on the analysis being performed. A “complete picture” in the truest sense is certainly a stretch, but it provides a suitable North Star for system and platform designers to work toward.
Existing hardware that is newly connected to the IoT offers one such opportunity to derive data from the physical world. Many digital devices already generate and store a wealth of information about their own operation and the environment they operate in. Connecting these devices such that it is easy for them to relay this data turns them into de-facto and potentially powerful sensors. In retail, these devices might include cash registers, electronic article surveillance (EAS) systems, HVAC control systems, and many others. By connecting your HVAC system to a platform that integrates many other sources of information about the store, for example, you’ve added another dimension to the view available to your analytics system – perhaps the temperature of the store has some influence on customer behavior that can be analyzed and understood in the context of other things that are happening in-store.
The IoT enables a whole host of new gadgets and apps, and they enhance user experiences in a variety of ways that I won’t go into here. However, as an ancillary benefit, these apps and devices deliver new opportunities to digitize information about the world that was previously unavailable. If the IoT enables store employees to carry devices containing inventory and product information for example, then those same devices might also be used as sources of data about where store employees spend their time and where they are while querying stock for customers. This sensor data can be just as valuable as the inventory needs the gadgets were designed to address in the first place.
When gathering data about events or elements, or adding dimensionality to existing data sets, we often find existing sensors to deploy and report back via IoT. However, if such sensors don’t exist, we now easily develop new sensors dedicated to our purpose. With so much connectivity so readily available, and with the costs of producing hardware even in small quantities continuing to freefall, engineers now produce and deploy dedicated sensors far more cost effectively than ever before, meaning that effectively nothing is out of the reach of our analytics capabilities. In retail, it means comprehensive analysis of inventory and product movement, shopper behaviors in-store and online, shopper/associate interactions, both in-store and online, and so much more.
The Internet of Things, while already huge and growing exponentially, has only begun to unlock many exciting areas of functionality for users of all kinds, from consumers wanting new kinds of experiences to executives hoping to understand complex business operations at scale. In the context of physical world understanding and analytics, specifically for brick-and-mortar retail stores, the IoT is a powerful source of structured data about the physical environment and the activities taking place therein. Existing devices that are now connected, new gadgets enabled by the IoT, and dedicated sensors that are now more feasible because of the IoT and other market forces can all be thought of as sensors providing data to our analytics systems. This is data that can be crunched, combined, and analyzed to provide real, actionable insights and optimization paths for businesses of all kinds.
Join the #IoT #RetailIoT and #inspiringretail conversations with George on Twitter @Kngpengwin & @RetailNext, as well as at www.facebook.com/retailnext.