Delivering the Right Retail Experience Requires an Edge Strategy That Connects to the Cloud

This post is sponsored by Intel? and based on the discussion in the Elevate the Edge podcast. However, Right-time Experiences is the intellectual property of Lopez Research and the opinions on retail are my own.

It’s clear that market expectations have changed for retailers in the past several years. Many retail experiences can’t fulfill today’s changing customer expectations. Retailers must reimagine and evolve digital and in-person experiences to create what Lopez Research calls right-time experiences. An RTE delivers the right information to the right person at the right time. These experiences are contextual, learning, predictive and prescriptive. To enable new experiences, retailers need to create a distributed infrastructure that supports analytics from the edge to the cloud. The architecture must be high-performing, resilient, and scalable – even at the edge.

In a recent Elevate the Edge podcast, with TIBCO Software Inc and Dell Technologies, we discussed how both companies are delivering new edge solutions for retailers with Intel?. This conversation covered the characteristics retailers should look for in solutions to create these right-time experiences at the edge.

Edge computing distributes computation and data storage closer to the source of where the data gets generated. Nick Underwood, Consulting CTO (Retail) at Dell Technologies, who spent 16 years at Target before joining Dell Technologies, shared how the edge is not new in retail, but edge technology is transforming the overall customer experience for retailers. He said, “Customers expect a seamless, frictionless experience both online and in-store. We often talk about the intelligent store, and that brings together many things.” Today the complexity and requirements of edge infrastructure have changed. Retailers must collect, analyze and act on data from a myriad of disparate devices in near real-time or real-time to support personalized retail experiences.

Retailers increasingly rely on applications that leverage machine learning and deep learning to deliver sales and support solutions that become learning, predictive and prescriptive. By rapidly processing large volumes of data, a retailer can improve flow, increase basket size and predict inventory requirements. Advanced software will use AI processing to provide prescriptive guidance for achieving the best outcomes by analyzing multiple scenarios.

Creating business agility with intentional architecture

Embracing edge computing with AI provides multiple benefits. The first is taking traditional use cases to an entirely new level. For example, computer vision solutions that couple streaming data analytics with real-time machine learning can enable a fast-food restaurant to understand how many cars are in the queue, how many occupants are in a vehicle, and if it’s a frequent customer. Leveraging other types of contexts, such as time of day and weather, can help a restaurant predict and prepare for shifts in food orders.

TIBCO noted the importance of processing multiple streams of data as one transaction and delivering AI inferencing in real-time to enable services such as self-checkout kiosks. To do this, you need to integrate and analyze data in real-time from multiple devices such as a scanner, the RFID mat, a scale, and a camera to support a single point of transaction. In a busy retail environment, the store will be processing information from multiple kiosks simultaneously.

Underwood picked up on this theme with a real-world example of how the right technology can assist in loss prevention. For example, a person may have two pounds of steak with a retail value of $40. But if the person wants to steal goods, they could scan a label that says it’s two pounds of carrots. When the customer places the item on the scale, the steak weighs two pounds, but they’ve just bought something that has a value of $4 instead of $40. The ability to leverage data from all the touchpoints at the self-checkout kiosk can help retailers identify this issue and notify a staff member. Self-checkout scenarios require advanced processing technology, such as sub-second processing of multiple data streams at the edge using artificial intelligence. The good news is that we have access to high-performance edge computing infrastructure today that enables a wide range of new retail use cases, such as augmented reality, computer vision, and virtual reality. TIBCO and Dell shared how new applications and use cases demand a processing architecture that supports inferencing at the edge, such as the 3rd Gen Intel? Xeon? Scalable Processor.

TIBCO described how access to real-time can provide agility and resiliency to support a company’s bottom line. During the pandemic, Panera Bread created a system called Data Pantry to understand the real-time inventory of perishable food. Armed with this information, Panera could sell excess inventory directly to consumers to minimize the losses.

Retailers need high-performance infrastructure at the edge to thrive.?When asked what retailers need to look for, Underwood said retailers need an intentional architecture that provides the potential to run applications or workloads in the right place and at the right time, which could be at the edge of the edge?in the case of point of sale. He shared how the traditional focus in retail was on security, resiliency, and availability. Underwood discussed how the data center and the retail store are rapidly becoming software-defined. Retailers need performance, insights and platforms that can provide flexibility with the performance from the edge through the cloud. Underwood said, “You need a platform for the edge, the core and the cloud that can deliver on that consistent, seamless experience and allow for rapid innovation, testing and learning.” This platform leverages a smart server-based architecture at the edge with a common control plane and data plane that gives you visibility.

TIBCO added that you need to pick platforms and frameworks that allow you to easily embrace a wide range of IoT devices today and the flexibility to support whatever comes next. TIBCO said you need to consider how much you can centralize the management of your edge ecosystem. The platform a retailer selects should also provide developers with a low code/no code experience.

He also mentioned how the edge is a team sport and that TIBCO’s partnership with Intel? allowed them to co-create platforms that analyze distributed data from a wide range of devices to drive real-time decision making

Meanwhile, TIBCO and Dell shared the importance of open source and secure systems. Both companies stressed that retailers need to understand the cost of placing data in a specific location in terms of dollars and data accessibility for analytics. Underwood said, “the cloud isn’t one destination.” Companies need information in public cloud hyperscaler systems and at the edge, but neither should it be an island. Companies eliminate any single point of failure by deploying an edge-to-cloud strategy.

?Conclusion

Best-in-class retailers collect large volumes of data, process it automatically, and use it to make decisions in real-time to support customer and employee needs. Innovative retailers aren’t just building to upgrade systems but are building infrastructure with a use case in mind. Meanwhile, TIBCO also stated that edge computing is about being nimble. He said, “Companies can leverage technology to be nimble and to respond to different conditions and situations, in whatever way is needed.” I couldn’t agree more.