Skip to main content


IoT Analytics: Smarter Retail with a TAP

The Internet of Things (IoT) gives retailers a fundamentally different view of the customer. By collecting information in real time, retailers can better understand what customers buy and how they shop—and can respond immediately.

Consider a customer who wants to buy a shirt but cannot find the right size because a shirt was misplaced. Traditionally, this results in a lost sale.

With the IoT, RFID tags can track a product's movement through the store. If an item is misplaced or sold, a salesperson can be alerted to replace the item.

This capability is particularly valuable for clothing retailers, who often have only a few items of each size on display but a full inventory in stock. Being able to locate inventory quickly increases customer satisfaction, eliminates missed sales, and enables more accurate inventory management and control.

Challenges to IoT Edge Analytics

Taking advantage of IoT edge analytics requires more than just connecting devices to the network. The truth is that adding IoT capabilities is far more complex than simply adding a new feature to an existing design. To be effective, IoT-enabled devices require an entire ecosystem to support them, from nodes to sensors to gateways to applications in the cloud. Typically, many types of devices are in use, so the ecosystem has to support a heterogeneous environment based on multiple connectivity technologies as well. Data needs to be managed, often remotely, and personnel need to be able to access data via mobile devices.

Business challenges also need to be considered. IoT is still an emerging technology, and it can be difficult to understand all of the options available. Development teams need a highly diverse skill set. The ecosystem needs to support change management, both to adapt products to how customer shopping patterns change as well as maintain device security. Retailers also need tight security to protect the data they collect and their customers' privacy. And perhaps most significant, the IoT changes how retailers operate at a fundamental level, driving change across the entire business.

Accelerating Design with IoT Frameworks

Developing a smart retail solution takes tremendous work and a large initial investment. For this reason, many OEMs are turning to companies like Altiux for design frameworks that can simplify the process.

Altiux provides a comprehensive suite of IoT middleware that serves as the underlying framework for IoT endpoints, IoT gateways, and cloud edge devices. Altiux middleware enables developers to build designs, take them to production, and scale them in the field. IoT capabilities can be introduced to existing product lines to convert them into connected – and smart – IoT-enabled products.

BoxPwR is Altiux's production-ready IoT middleware for IoT devices, from extremely low-power and processor-constrained nodes to full-service gateways (Figure 1). With its modular, configurable, and standards-based approach, it provides the connectivity, security, and value-added features essential for IoT deployments.

Figure 1. Altiux BoxPwr device middleware supports a range of connectivity protocols and device types to accelerate connected retail applications. (Source: Rakuten Infoseek

BoxPwR supports Intel® Quark, Intel Atom®, and Intel® Core processors. This gives developers the flexibility they need to address varying security, reliability, and performance requirements.

Altiux also provides a device management framework that provides a standards-based approach to heterogeneous IoT ecosystems. Framework features include provisioning, registration, commissioning, and firmware over the air (FOTA) updates, as well as several security modes.

The multi-cloud framework greatly simplifies cloud deployment (Figure 2). Because this framework enables seamless data routing to any cloud – even multiple clouds simultaneously – developers don't have to decide on a particular cloud until they are ready.

Figure 2. Altiux device management framework supports multiple clouds and connectivity types to ease the IoT cloud platform selection process. (Source: Altiux)

The Trusted Analytics Platform

The Intel® Trusted Analytics Platform (TAP) is an extensible open-source platform initially developed by Intel and available to the IoT development community. It further accelerates the development of IoT-based ecosystems.

Part of the value of Intel® TAP is to reduce the amount of data transmitted to the cloud. For example, using algorithms that are part of Intel® TAP, gateways collecting RFID tag data can determine that tags that have not moved and do not need their position updated. Thus, only changes are sent to the cloud.

The results can be significant: Rather than sending thousands of updates per second to the cloud, only the handful that matter are sent. Intel® TAP implements the right level of intelligence in gateways to keep sensors simple and cost-effective. At the same time, it optimizes use of cloud resources and network bandwidth.

The information collected through Intel® TAP can be used in many ways. For example, at the end of the day, all RFIDs that are not in their correct location can flag an alert so that they can be collected and put back in the right place for tomorrow's customers.

Intel® TAP also supports the use of rules to prioritize actions and define more complex automated behaviors. For example, after accessing inventory records, Intel® TAP can ensure that product that is low in inventory is collected and put back in the right place first. The purchasing department can then be alerted to buy more of this item.

Data can be used to identify trends, such as whether some items are more likely than others to be misplaced. This can help sales personnel stay on top of these items. This data can also be used to help assess why these items are being selected but then put back and not purchased. Retailers can then begin to understand the reason why some products sell well and others don't so they can adjust product placement, displays, and pricing to improve sell-through for these items.

Improving the Retail Experience and ROI

IoT edge analytics is the foundation for smart retail, and enables retailers to create a better in-store experience for their customers while increasing their profits. With the right hardware and IoT-based middleware, developers can accelerate the design of these powerful smart retail IoT ecosystems.

Altiux pre-built frameworks are already ported to Intel hardware and provide a solid and robust foundation to begin design. Because these frameworks are production-ready, they substantially reduce design complexity and speed delivery of product to the market.

Since the frameworks are built on open standards, developers can take advantage of new IoT standards, protocols, and technology as they reach the market. In this way, developers – and their customers – can have confidence that the products they design today will continue to operate tomorrow.

About the Author

Nicholas Cravotta is a veteran of the electronics industry. He has been technical editor for EDN, Embedded Systems Programming, and Communications Systems Design, and was the founding editor-in-chief of Multimedia Systems Design. During his years as an engineer, he designed hard real-time embedded systems, wrote application software for PCs and workstations, built an operating system from the ground up, and developed in-house software and hardware development and test tools, among many other projects. He has written over 600 published articles and has taught programming and technical writing at UC Berkeley. When he isn’t writing about engineering, he is an award-winning game designer for BlueMatter Games where he focuses on innovative ways to engage people, including the home-version of Escape the Room and Houdini, the reconfigurable disentanglement puzzle. He was recently a contestant on the reality TV show, “The Toy Box” showing the Pinata Backpack.

Profile Photo of Nicholas Cravotta