Vashisht Madhavan & Sean Loughlin


Much has changed in the last 20 years with the onset of the technological age. Households and businesses alike rely on innumerable electrical devices to operate efficiently. However, the increase of electrical devices has brought not only an increase of energy usage but also an increase of wasted energy. It is time to bring energy management into the 21st century. The problem we are addressing is the abundance of wasted energy that goes on every day in households. By combining smart-sensors to monitor and control lights, appliances and outlets and location based tracking of the user we will provide an innovative solution to allow consumers to more effectively control their energy usage.


With the current state of household usage tracking and smart metering, PG&E only provides information about how much energy the overall household uses and how that usage fluctuates between each day of the month. An example is given below:

external image CHymSxnHWfl-y6iuntw58OBIoX25_YhjzhE1WhzDT2SzsOETqQbN20yiG0gDQyfPreooubyrJ1CbCQWWJvscPfJeLCyZoa11511LA1alziuWLfOrioc48qqmVtL1y-biusmlB0Uf
Screen Shot 2015-12-04 at 11.33.19 AM.png

Although this does give users insight into which days they are consuming more electricity, many of these charts are provided at the end of the month, at which point users forget what they used electricity for.

We hypothesize that by tracking information about which parts of a household are using more electricity than others and providing this information more frequently, PG&E can provide customers with more effective feedback about their energy usage. With this feedback, users can cut costs and save energy that would otherwise be wasted. PG&E can also use this information to inform demand response and distribute energy more intelligently.


To address the lack of usage statistics tracked by PG&E, our solution proposes to track electricity usage through outlets within a user’s home. By installing sensors in home outlets, PG&E can get a sense of the layout of a given household and track which rooms in the household use the most electricity. In addition, PG&E can track GPS signals from the sensors to figure out which room a user is currently in.
After aggregating this sensor data, PG&E can provide an online platform for a user to view usage patterns throughout the day and understand which appliances are causing spikes in usage. The platform would need to be accessible(i.e. a mobile app) and arrange usage data in a way that is easily understood.

external image bu3jNkETgzEdF_qW_9T0oRWtzQez19ZscwsJJu6ob_TBgAgeR6iDY6VKSotjyEvnMmexNbnwBWfxGB5hSdf22UEamEwR-g3aWrXPOoljceV-xug1F-RaY1l-IptK8B1t897LA6zy

A snapshot of we envision the dashboard to look like

Measuring Experiment Success

To know whether this solution would be effective, we would need to understand how users would change their usage patterns with this advanced tracking. We would also need to understand how much value this information would bring PG&E.
We could start off by selecting test users who would have sensors installed in one outlet in every major room in their household. After this, users would create online portals with all their relevant usage statistics. At this point, PG&E can track how often users log in to their online portals and request periodic user surveys to understand what users think of the product and how it has affected their energy usage.

external image sHc97G1vTsIWDbaejd31Rd5uHQd4JJKJbrRcfhvWFIt9Mon4dTtLEiobuL3wLuQ2Meew_s26T4u8GCoLnoKmeMeN-AXbPraIvE_7Qxali7cJ0GP2yZ3_B5BiddaGDaVzh2amLikY

We would try to create a plot like this to measure the actual impact of these smart sensors.


Based on the results from the measurements described above, PG&E can assess what value the information gives them and see if outlet sensors do change the way people use energy. If the experience with test users show positive results, PG&E can go about trying to implement this at scale. One thing that may hinder further developments is the high startup cost to make accurate sensors and install them in households. Although users may care about energy savings, they may not want to pay for these kinds of metering tools or go through the hassle of installing these sensors in their homes. Another bias to consider is that the users most likely to install these sensors are those who already take active steps to save energy, which could possibly skew the data gathered by PG&E.

Data Sources

The majority of the data that our solution uses comes from sensors in household outlets. We would obtain this information by allowing the device to send packets to PG&E servers that would store the information. Here is a list of things we plan on tracking, along with information about how we would obtain it:

  • time - date, day, hour,etc.

    • attach timestamps to the packets sen

  • location(long,lat) - sensor GPS location

    • give sensor GPS functionality

  • usage (kwH) - how much electricity is being used for the outlet

    • send peak power over the timeframe between packets

  • appliance - which appliance is using the outlet(i.e. fridge)

    • a user would enter this information in the online portal. They would enter which appliance corresponds to a given sensor id

  • sensor id - identifier to differentiate sensors in the household

    • preset in sensors

  • other customer data - compare the user’s data against that that of other users

    • They can compare their individual energy consumption against that of the neighborhood and measure specific appliance consumption against others that have the same appliance (among other metrics) to maximize efficiency.

Improving the User Experience with Data

Right now, the current paradigm in the consumer energy industry doesn't give the customer enough information to actively manage and generate insight from energy reports. At best, smart meters let consumers know which days of the month and during what hours they use peak energy. However, by providing more precise metrics, the new data will improve end user experience by allowing the customer to more effectively manage their home energy consumption. By installing sensors to more precisely track energy usage and then putting this data on an online platform, the consumer will be able to learn more about their energy usage and make adjustments to optimize the efficiency. For example, consider the user generates his new detailed report and finds out that his air conditioning bill costs 3x as much as the neighborhood average. The report would highlight this discrepancy and would the necessary information to conclude that he either needs to improve his insulation or invest in a more efficient air conditioner. Additionally, as the new smart system, using GPS real-time tracking, better learns the customers behaviour, the system can provide assisted and independent of the house. For example, if the oven is left on, and no one is in the house, the system can send the customer an alert asking if he wants the system to turn off the oven. Or if during a time the system knows the customer is at work, the system can shut off the bathroom light that the customer left on.