Skip to main content
Skip to footer
Contact Us
We are taking you to another website now.
January 10, 2019

As the number of retailers have increased over the years, so have their challenges. Customers today are more savvy and not tied to any brand, they prefer seamless and multichannel buying experiences. They demand instant gratification, and superior experiences have made it imperative for the retail industry to be constantly evolving.

Customers want to move across online and offline experiences, where they can view, touch and feel a product in-store, and explore similar products, visualize how it looks on them and make the purchase online. Retailers who capture customers’ purchases and preferences across their online and offline activities stand a better chance to create personalized experiences.

To do so, the store needs to gather historical data such as purchase history, products viewed on the website and consumer data, analyse it and process it in real time to be able to recommend the right products. Facial recognition or other new in-store experiences like beacons – a small Bluetooth low energy (BLE) device – can detect as soon as a customer enters a store and send targeted content based on their interests to their smartphones. In many cases, the amount of data would be too huge, expensive and time consuming to be sent, processed and stored at a central location. This is where edge computing proves to be an ideal solution.

Retail stock inventory is another area that can benefit greatly from the use of edge devices and edge analytics. Edge devices such as sensors, RFIDs and cameras can make it easier to track inventory, their exact location in the store, items that are fast selling and in demand, and items that are misplaced. This helps the store clerks and managers keep their store well stocked and use this data to generate analytics to understand sales trends, customer preferences, optimize shelf space, and automate the reordering and receiving process. Edge devices placed close to the billing area can also be used to gather customer feedback on their overall experience, convert speech to text, analyse their sentiments and use it to improve the store experience.

The shopping experience of the future also involves leveraging AI and gesture-based computing to help customers try out the outfits virtually. Robot assistants with inbuilt 3D scanners and image recognition technologies to guide customers  to products, assist in supply chain management by checking stock, fix pricing errors and detect misplaced items and smart displays, are but a few indicators of how the retail experience is about to be transformed. All these involve large amounts of data being collected, and not all of these are relevant to be sent to the cloud or stored for long-term. Naturally, this would take up an extremely high bandwidth in a centralized cloud network, given that thousands, if not millions of sensors will need to consistently stream data to the cloud. This is where edge computing comes into play. The edge networks would help provide real time insights and save costs of transmitting entire data to the central cloud and computing resources involved.

Additionally, virtual and augmented reality have made their presence felt in the retail industry. According to a research by Retail Perceptions, around 61% consumers prefer stores that offer AR experiences — and 40% of them would pay more for the product if they have the chance to experience it through AR. However, VR and AR experiences can be resource intensive and need fast response times as even 15ms delay can cause motion sickness in the viewer. Edge computing prevents this by providing the underlying hardware to power high-end VR and AR experiences by utilising the ‘edge cloud’, closer to the end device, thus allowing high-end graphical performance with low latency.

Edge computing introduces new business opportunities and experiences, but comes with its own challenges of remote site management, security risks, seamless integration, additional investment and dealing with a fragmented vendor market. Edge computing technology is still nascent and the evolving market needs to mature to clearly understand use cases that can benefit from edge processing. The right selection of sensors and edge gateways, deployment of application workloads on edge nodes such as gateways or micro data centers, partitioning of applications at the edge, where and when to use edge nodes and implement real time data processing at the edge are few of the many considerations. Edge computing is typically implemented on several nodes that lie between the edge device and the cloud; including access points, base stations, gateways, traffic aggregation points, routers, and switches.  However, these solutions may not yet be suitable for handling edge-related workloads. Vendors have started developing software to cater to edge computing, but testing and validation is still necessary to real world use cases. Discovery mechanisms to find specific nodes that can be leveraged in a decentralized cloud set up are necessary and current methods used in the cloud are not practical.

Edge orchestration is critical, owing to the huge number of devices involved, their heterogeneity and data formats, unreliable network availability, need for automation of workflows, fault detection and self-healing. It is crucial not to overload the edge nodes with computationally intensive workloads so that their intended workloads and user service levels are not impacted.  Task partitioning and scheduling should be done based on edge nodes peak hours of usage. Gartner predicts that by 2020, more than 25% of identified attacks in enterprises will involve IoT. Edge or embedded devices have size, weight and power constraints that make an on-device agent impractical. There is no homogeneity of operating systems that we have in traditional end-points; the things we carry are easy to hack and there is no common interface framework to provide warnings. Hence, a security framework with tools need to be developed and evaluated. Considering these challenges, it is important to have a partner with deep expertise in digital and niche technologies such as edge platform development and integration, sensorization, gateway platforms, cloud computing, data center technologies, edge cyber security, edge analytics and predictive maintenance to help clients address the risks involved and have a successful implementation.

Retail shopping will change more in the next 10 years than it has in the last 1000 years, and the shopping of the future will see the physical and the digital worlds converging to provide the combined experience of the tangible, authentic brick and mortar store and the high-tech features of online shopping. With the right architecture and technologies powered by Edge, IOT, Artificial intelligence and Virtual reality, Retail industry can deliver along the key principles of enhanced customer experience, optimizing in-store experiences and driving new insights based on data gathered within the store, which will enable them to attract loyalty and build the brand.


Lekha is an Enterprise Architect with two decades of IT experience, and leads CBO - Edge computing Center of Excellence. Her current role involves providing consulting services to customers in edge computing, defining end-to-end edge architecture, RoI calculation and blueprint formulation. She is also involved in building solution offerings in edge computing, understanding the market needs, interfacing with vendors in the areas of edge computing and IoT, and building industry specific solutions for customers. She is a certified AWS cloud solutions architect, TOGAF and AIIM information governance architect.



Thank you for downloading

Your opinion counts! Let us know what you think by choosing one option below.