By leveraging the power of IoT-enabled data analytics, retailers have new opportunities to drive operational efficiencies and grow profit margins.
Making the transition from manual processes to digital automation is becoming a top priority in many retail business models, including grocery and convenience stores. From labor and technician shortages to consumer and competitive pressures and shifting market landscapes, retailers face a challenging array of dynamics that demand efficiency improvements in nearly all aspects of their operations. By leveraging the power of internet of things (IoT)-enabled data analytics, retailers have new opportunities to drive operational efficiencies and grow profit margins.
Most retail stores are data-rich environments from which information is generated continually among a variety of sources on “the edge,” such as refrigeration and HVAC equipment, lighting systems, food and beverage stations, and point-of-sale (POS) transactions. Transforming this raw data into operational insights requires a technological infrastructure that connects IoT-enabled assets (on the edge) to aggregate data in a secure cloud-based platform. Once there, data can be harvested by data scientists and processed via advanced and ever-evolving machine-learning (ML) algorithms and/or artificial intelligence (AI) techniques.
This “insights from the edge” approach represents a paradigm shift for retailers that typically access and interpret their operational data within the scope of asset-level control devices. It also greatly expands the potential power of data into a robust analytics environment that delivers continuous insights.
Understandably, busy store operators are often dubious about the prospect of dealing with more data. They know what it’s like to be inundated with too many alarms without being guided on which issues take priority or what actions are needed to resolve them. That’s one key reason that it’s critically important to work with a data analytics service provider that understands challenges and has the broad refrigeration domain expertise needed to bring only essential operational insights to the surface.
Accelerating Insights, Locally and Enterprise-Wide
Providing cold food and beverages, hot coffee, and fresh food offerings to consumers on the go is at the core of any retail brand promise. If refrigerated food cases and beverage coolers don’t maintain their optimal temperature setpoints, profit margins and customer loyalty may suffer. At a store level, operators need to know immediately when temperatures drift from these setpoints, so that they can take appropriate actions.
At an enterprise level, administrators need tools to help them take a holistic view of their entire fleet, uncover opportunities for large-scale improvements and optimize operational expenses. As many organizations deal with labor shortages, they also need smart tools that help their limited resources focus their efforts and address the most pressing issues at hand. Edge-derived data insights can achieve these objectives quickly and effectively.
One of the key differences between edge-based analytics and traditional asset-level manual interpretation is the sheer speed of generating insights. In a traditional model, an alarm would indicate when a cooler temperature deviated from its optimal setpoint. Once an alarm came to the store operator’s attention, a technician would be dispatched to troubleshoot and conduct an analysis. Then issue resolution would be dependent on the information available to the technician at the time of troubleshooting.
In contrast, an edge-based data analytics model would enable ML models to assess and diagnose issues rapidly. By consulting with store operators, data experts can build custom ML algorithms that address specific areas of concern — not within months or years, but within days or weeks. They can enable automatic health checks as necessary to make sure systems are performing as expected and parameters are set properly. If not, they can recommend changes to correct issues so key assets (e.g., meat cases) can be restored to optimum performance, minimizing the need for truck rolls or technician intervention while reducing energy consumption.
As is often the case, an alarm or issue in one store location is an indication of a trend occurring in other stores. Data analytics teams can extend their remote health checks to evaluate cooler and freezer performance within the larger enterprise store network. By leveraging this robust data infrastructure, new ML algorithms can be created and deployed easily to drive continuous improvement across the entire enterprise.
Adapting to an Emerging Electric-Vehicle Landscape
The retail market landscape is continuously evolving to accommodate consumer demands. To differentiate, many companies are expanding on their fresh food offerings and providing higher-quality meals for discriminating consumers. Also, as the transition to electric vehicles (EVs) progresses over the next several years, retailers are planning for a future in which consumers will need them to provide EV-charging stations in addition to traditional offerings.
At the same time, many retailers are more closely monitoring their carbon emissions and exploring ways to optimize the energy efficiency of their stores by participating in grid-interactive arrangements with their local utility providers.
Balancing traditional core business and expanded food offerings with emerging EV-charging stations and sustainability initiatives will introduce a new set of challenges. IoT-enabled data analytics — combined with the supervisory control from an energy management system (EMS) — can provide the automated tools and operational insights retail store operators need to help them integrate EV stations and achieve energy efficiency goals, all without compromising food quality/safety or decreasing product shelf life.
While an EMS is critical for managing refrigeration, HVAC, lighting and potentially EV-charging assets, it also serves as a gateway for communication with the cloud. By merging EMS data streams with other edge devices, analytics teams can create ML algorithms that take all variables into consideration — such as refrigeration load requirements, utility rates, grid availability and weather data — to make the most optimal, automated and informed decisions.
Consider a potential future scenario. Multiple cars pull into a store EV-charging station during the heat of the day when the refrigeration load is at its peak. The local utility has asked commercial building operators to prepare for the possibility of brownouts. The question becomes: How can a retail store operator make resource allocation decisions that provide the maximum benefit to customers, their operations and the grid? Automated algorithm-driven intelligence will be essential for adapting to these likely scenarios.
Future Potential: Combining Sales With Operational Data
Today, operational and sales data exist in separate silos. But what if retail store operators could marry their data streams to leverage food supply and demand trends to their advantage? As IoT-enabled, edge-based technologies evolve, operators will undoubtedly uncover new opportunities to align operational and sales data to create promotions, influence consumer behaviors and grow profits.
For example, if a perishable food item is nearing the end of its usable shelf life, operators could run a new sales campaign to move the product at a discounted rate — a scenario that benefits both the operator and consumers. By leveraging product inventory data, operators could effectively plan and coordinate sales cycles and better optimize inventory supplies with consumer demand.
IoT-driven insights could also help retailers optimize other key areas of their operations — such as fresh food preparation or even coffee-brewing cycles — that currently rely on the availability of store personnel. Data insights could suggest how to optimize brew cycles by identifying when a batch is no longer fresh or when is the most optimal time to brew a new one, or determine how long the current batch of hot dogs has been on the rollers. Improved data analytics could potentially provide these answers.
By extracting insights from this sales-related data, store operators could grow sales and gain consumer traction in an increasingly competitive market space. These are just a few examples of how retailers can drive process improvements by transitioning from manual monitoring to data-driven automation.
About the Author
Paul Fullenkamp is manager, cold chain data science at St. Louis-based Copeland (formerly Emerson Commercial & Residential Solutions). Fullenkamp has 18 years of engineering experience at the company. He earned a bachelor’s degree in mechanical engineering from the University of Toledo and a master’s degree in mechanical engineering from the University of Dayton.