Utilities worldwide are installing smart meters on homes and businesses, which means there could be as much as 50 terabytes of energy data that can emerge from a million or so homes in a year. The problem has been that there haven’t been very many ways to make good use of all this data to benefit the average consumer. But a startup called Bidgely, which raised a series A round from Khosla Ventures, says it has created algorithms that can dig into real-time smart meter energy-consumption data, can reduce consumers’ home energy use by between 4 percent to 12 percent, and can also deliver other beneficial home services to consumers.
While other companies make similar claims, Bidgely’s big selling point is that the company says it can take smart meter data, utility data or energy data from a Zigbee-based router in the home, and be able to tell which appliances are consuming what amount of power in a home in real time without having extra hardware or sensors on each plug or appliance. Other companies that focus on this — which people in the industry call appliance-level “energy disaggregation” — need the consumer to buy smart plugs and stick them on every appliance and outlet. That makes these type of hardware-intensive options basically a non-starter for average consumers.
Bidgely says it is the first company out there that can detect appliance consumption patterns without additional hardware and sensors. For example, if the freezer is left open, or a pool pump is old and not running efficiently, it can determine that with its algorithms, explained Bidgely’s co-founder and CEO Abhay Gupta. Gupta previously worked at smart grid company Grid Net, energy company Echelon, and Sun Microsystems.
Gupta said Bidgely has been able to create these types of algorithms first partly because it was able to obtain a significant number of Google Power Meter users after Google shut down its Power Meter application in the summer of 2011. Some of these Power Meter users were using TED real time energy gadgets, and Google Power Meter had its own algorithms to discover when appliance were being used, so Bidgely was able to spend over a year crunching these detailed data sets and creating its own algorithms.
Bidgely gets its data from a few sources, each of which have their own limitations. Gupta told me that through the 8 years that he’s spent working on energy data companies, he knew what was available and wasn’t available from different data sets.
Utilities in some states are working on a Department of Energy-developed initiative called Green Button, which opens up smart meter energy consumption data to third party developers like Bidgely. That data is generally energy consumption data from 15-minute or 1-hour intervals. The problem though is that not all of the utilities in the U.S. are aggressively embracing the Green Button initiative (though California’s utilities are). Also, Green Button data is not always real time and can be a day-old — that makes it far less useful.
If a consumer is one of the few that has bought a router or gateway that connects to the smart meter via a Zigbee wireless connection, then Bidgely can use the data off the router or gateway in real time. Bidgely is working with hardware partners like Digi and Rainforest for this type of hardware (see gadget to the right). Utilities also have these types of Zigbee-connected device projects in the works, and a small number of them are starting to open up that Zigbee energy data to third parties.
Finally, utilities that aren’t opening up their energy data via Green Button or Zigbee device projects can connect with Bidgely directly. But then Bidgely has to connect with the individual utilities and convince them to trust Bidgely with their customers’ data. It’s not an impossible task, and Gupta said it has three trials that will start soon with utilities.
One reason utilities might want to work with Bidgely is that utilities are looking for ways that consumers can benefit from smart meters, as there’s been some backlash from consumer advocate groups — and consumers themselves — that smart meters are only benefiting the utilities. Startup Opower has managed to win over utility customers for its first generation of energy billing services.
But Opower’s service can reduce consumer energy consumption by 1.5 percent to 2 percent — compared with other services like Bidgely that are aiming for far higher reductions. Aggressive utilities are looking to move to the next-level of energy efficiency services.
Once Bidgely gets the energy data it needs, it has created a recommendation engine that uses the data to give its customers real-time advice and instructions for how to reduce their energy consumption. These could be energy-saving techniques — like get a new pool pump or close your fridge — but they could also be non-energy specific services about the home. For example, closing your fridge can be helpful just to save the food inside from thawing.
The recommendation engine uses cloud-based large machine learning techniques to deliver personalized recommendations, and also incorporates more big data sets into the engine like social, weather, and demographics. Bidgely uses No-SQL database techniques.
Other companies that are relying on algorithms, and big data to reduce consumer energy consumption include Nest Labs, C3, Opower, Silver Spring Networks, EnergyHub, Ecofactor and more.