Are Smart Meters Too Smart?

Photo credit: iStockphoto/ValeryBrozhinsky

More than AUD 2 billion has been spent on the installation of electricity smart meters in the Australian State of Victoria. According to a recent report by the auditor general, they have delivered few, if any, of their intended benefits.

The smart meters, installed initially to monitor and control energy use and drive efficiencies, are about to find a further use with a landmark deal for AWS to harvest the data they collect.

AWS has joined New Zealand utility company Vector in a shared partnership, which will see data from the smart meters collected and then analyzed.

Data from 1.6 million internet-connected smart meters, mainly in Melbourne but in other parts of Australia and New Zealand, will deliver information on consumers’ consumption and the performance of their networks.

The project will allow power companies to “develop tailored product and pricing solutions for their customers based on their energy consumption habits,” according to the announcement.

Beyond that, the use of the data is more wide-ranging and full of fascinating potential.

Personal habits captured

The smart meters reveal information such as the number of televisions and other appliances in the home, the age and performance of the fridge, and also the time when people are at home using power, how many people live there, and when the house is vacant.

This data has been likened to the purchase of all of a city’s CCTV footage, with critics saying that — even though the data is anonymized — it is nevertheless an intrusion into people’s privacy.

Unsurprisingly, Vector and AWS disagree, with Vector chief executive Simon Mackenzie describing it as a “really exciting opportunity to unleash the power of data to make customers products and services and costs much more efficient.”

“No-one wants to sit and analyze every five minutes of their energy supply,” Mackenzie said.

“This will ultimately involve solutions that, through machine learning or other types of technologies, enables customers to opt-in to products or services and have it managed in a smart way, privately behind the scenes to have a better experience and lower cost.”

He envisages a future where the data can be used for third party providers to enter the market and target offers such as home charging devices for electric vehicles to relevant consumers.

The AWS managing director for the commercial sector in New Zealand, Nick Walton, says the deal can help drive the digital transformation of the energy sector.

“We think that we can really unlock a lot of the value from this data and start to provide solutions back to customers with more choice and more visibility into their power consumption patterns,” he said

“Then, consumers can take action on choosing to, say, charge things off-peak, or just get greater visibility to make smarter decisions that suit themselves and their lifestyle.”

Where are the tangibles?

Intrinsic in the deal, however, is a recognition that the smart meter rollout — particularly in Victoria where it has been most ubiquitous — has failed to deliver anything tangible so far.

Only through analyzing the data further and interrogating it with machine learning can real value be delivered. The AWS deal is an additional layer of data collection and analysis which are needed to provide any significant value from the project.

The Victorian program has been going for a decade now, and more than AUD 2 billion has been spent. But according to the auditor general, the state government department responsible “does not have a good understanding of the cost of the program, which it does not track.”

Consumers would also appear to be confused or indifferent, and two-thirds of Victorians did not understand the benefits of the smart meters.

The Victorian auditor general’s report found that only 0.27% of consumers had subscribed to flexible pricing offers associated with smart meters, well below the target of 4% by 2014 and 15% by 2017.

Other critics say the only benefit so far has been to the power companies, who have been able to save money by sacking the meter readers who used to roam the suburbs reading people’s old-fashioned meters so their bills could be calculated.

Both AWS and Vector say the data won’t be linked back to individual consumers, and Amazon won’t be involved at all because it is a separate company from AWS.

Vector is also audited by Australian energy regulator AEMO as an accredited data provider.

These protections do not satisfy data advocates, who describe the deal as opaque because it is not clear exactly what data is being collected, who it is shared with, and what it is being used for.

They also point out that such a deal would be impossible in Estonia, the only other country to roll out smart meters en masse. There, third parties are denied access to smart meter data if users have granted access to a specific service provider.

The uneasiness in data correlation

One of the significant trends in the data industry is aggregation, as data silos are broken down and analysts – both humans and machines – join the dots and find new insights.

Aggregating data from the smart meters with credit card and even traffic data from the same area could deliver an unprecedented view of human activity and drive business models and innovations which are yet to be even dreamt of.

That sounds exciting, but the reality is that at this point, we don’t know if the price of greenlighting these data innovations is going to be worth it for consumers and if the balance with business use is going to be a fair one.

In the meantime, AWS and Vector are set to begin their project later this year.

Photo credit: iStockphoto/ValeryBrozhinsky