THE OPPORTUNITIES OF TECHNOLOGY DATA AND AI

Automated intelligent energy management could be the answer – but many retailers are struggling with data, asset connectivity and system integration. Where is the sweet spot?

A SHARED FUTURE VISION EMERGED IN OUR INTERVIEWS

Machine learning enables automated optimisation of energy use and maintenance actions.

Many retailers have aspirations to leverage advanced digital technology to better manage their estate, releasing energy savings and maximizing the return on investment in new energy assets. A shared future vision emerged from our interviews.

An AI, machine learning ‘brain’ that can read data from multiple sources (occupancy, weather forecasts, performance of fridges and HVAC systems, temperature, customer and employee comfort levels and energy prices) and automatically adjust settings for heating, lighting, fridges and chillers, while continually learning how to optimise for comfort, asset performance and cost.

Controls would extend to manage on-site energy generation from solar panels, determining when to charge or dispatch batteries and responding to signals from National Grid for demand-side power adjustments.

Digital twins are also seen as leading practice in running scenarios to understand how adjusting device settings would impact temperature, performance and costs.

ONGOING STRUGGLES

In reality, these advancements feel far away, with many retailers struggling with some of the basics.

Retailers struggle with limited connectivity to remotely monitor and control energy-hungry assets such as fridges, chillers or air handling units.
Data is fragmented across unconnected systems; meter data is received with a 24-hour lag in batches from energy management systems, while asset maintenance systems highlight potential performance issues with fridges but without a connection to energy consumption and cost.
Manual analysis is commonly required, stitching together multiple data sources to understand energy costs and areas for savings across different assets. Engineers are often dispatched to check if fridges are failing, incurring avoidable costs when issues could be resolved by remote adjustments to settings, or with more granular data on performance (e.g. detailed fridge compressor telematics).

Legacy Building Management Systems are limited in functionality, lacking intelligent capability, either needing replacing or requiring bolt-on systems for effective asset management. There is a lack of accurate asset registry data to understand what devices and assets are in stores, making it difficult to know which ones have the potential to be remotely controlled or transmit performance data.
There are often multiple BMS and connectivity systems across the store portfolio, creating a challenge to move to a central system for monitoring and control.
Existing data integrity challenges mean the reliability of data from devices may not be sufficient to feed into automated systems.
Systems are often not integrated to allow end-to-end automation of energy and device management. For example, CAFM maintenance systems may not be connected with device performance, BMS or energy management systems.