As I have mentioned here previously, Smart Meters are going to bring a flood of data to utility companies which will need to be properly managed and which can be a source of intelligence for the utility, if they mine it well.
At the recent SAP Sapphire Now conference, UK Energy retailer Centrica showcased their Smart Meter Analytics application running on SAP’s HANA. HANA is SAP’s in-memory computing solution (In-memory computing moves data off traditional storage on servers and into RAM, providing a performance boost over having to read the data off disks).
Centrica are the largest single instance utility on an SAP system with 18 million residential accounts for and one million business accounts. Right now they are billing residential accounts every three months and they are managing 75 million meter reads per annum.
With the move to smart meters, Centrica will take electricity reads every 30 minutes and gas reads once per day. This means a shift from 75 million meter reads per annum to 120 billion meter reads a year. 120 billion – that’s billion with a b. That’s a phenomenal amount of data to have to deal with. Doing any kind of traditional analytics on a data set that large would very quickly get totally bogged down. One of the interesting things about HANA however is that the performance scales linearly with the hardware. If it’s starting to slow a bit with 120 billion meter reads, throw a couple of extra terabytes of ram and servers at it and hey presto you are back in business, or so the theory goes!
In the demo above, Centrica are using the analytics to examine their customer segmentation. They can look at the energy profile of similar businesses in a specific area and where there are anomalies, they can work with those businesses to help them cut their energy consumption until it is more in line with their peers. Increasing pressures to be energy-efficient and to reduce carbon footprints are being looked on by Centrica as an opportunity to open an energy services business, going to customers to help them to become more energy-efficient. The Smart Meter Analytics application is going to be crucial for this new practice within Centrica.
In the residential market the Smart Meter Analytics app presents householders with an Energy Efficiency Scorecard. This facilitates charting actual usage patterns against the various tariffs Centrica offers to find the optimal rate plan. The score card also allows home-owners to compare their energy efficiency with similar households in the same area.
This is an impressive use of Smart Meter analytics and it presents hugely useful information but to avoid a short MTKD (Mean Time To Kitchen Drawer – the time it takes for people to get bored with this app and metaphorically stuff it in the kitchen drawer), the scorecard has to keep homeowners engaged. There are several ways of doing this. One of main ones would be to bring in some gamification – make a game of it. The Scorecard could set energy reduction targets for people to meet in the next set period, allot points or awards for achieving reduction targets, build social media in for sharing achievements and put up a leaderboard to add some real competition to the game.
Energy management applications to-date have suffered from the fact that they were working with estimated data a lot of the time, from lack of a significant networked dataset and from a lack of a usable analytics engine. Large energy retailers, like Centrica, have the opportunity now to change that and to make energy management in the home compelling. Let’s hope they make it so.
Disclosure – SAP paid my travel and expenses to attend Sapphire Now.