Our team wanted to make a smart power bar device to tackle the challenge of phantom power consumption. Phantom power is the power consumed by devices when they are plugged in and idle, accounting for approximately 10% of a home’s power consumption.  The best solution for this so far has been for users to unplug their devices after use. However, this method is extremely inconvenient for the consumer as there can be innumerable household devices that require being unplugged, such as charging devices for phones, laptops, vacuums, as well as TV’s, monitors, and kitchen appliances.  We wanted to make a device that optimized convenience for the user while increasing electrical savings and reducing energy consumption.
What It Does
The device monitors power consumption and based on continual readings automatically shuts off power to idle devices. In addition to reducing phantom power consumption, the smart power bar monitors real-time energy consumption and provides graphical analytics to the user through MongoDB. The user is sent weekly power consumption update-emails, and notifications whenever the power is shut off to the smart power bar. It also has built-in safety features, to automatically cut power when devices draw a dangerous amount of current, or a manual emergency shut off button should the user determine their power consumption is too high.
How We Built It
We developed a device using an alternating current sensor wired in series with the hot terminal of a power cable. The sensor converts AC current readings into 5V logic that can be read by an Arduino to measure both effective current and voltage. In addition, a relay is also wired in series with the hot terminal, which can be controlled by the Arduino’s 5V logic. This allows for both the automatic and manual control of the circuit, to automatically control power consumption based on predefined thresholds, or to turn on or off the circuit if the user believes the power consumption to be too high. In addition to the product’s controls, the Arduino microcontroller is connected to the Qualcomm 410C DragonBoard, where we used Python to push data sensor data to MongoDB, which updates trends in real-time for the user to see. In addition, we also send the user email updates through Python with the time-stamps based on when the power bar is shut off. This adds an extended layer of user engagement and notification to ensure they are aware of the system’s status at critical events.
Challenges We Ran Into
One of our major struggles was with operating and connecting the DragonBoard, such as setting up connection and recognition of the monitor to be able to program and install packages on the DragonBoard. In addition, connecting to the shell was difficult, as well as any interfacing in general with peripherals was difficult and not necessarily straightforward, though we did find solutions to all of our problems.
We struggled with establishing a two-way connection between the Arduino and the DragonBoard, due to the Arduino microntrontroller shield that was supplied with the kit. Due to unknown hardware or communication problems between the Arduino shield and DragonBoard, the DragonBoard would continually shut off, making troubleshooting and integration between the hardware and software impossible.
Another challenge was tuning and compensating for error in the AC sensor module, as due to lack of access to a multimeter or an oscilloscope for most of our build, it was difficult to pinpoint exactly what the characteristic of the AC current sinusoids we were measuring. For context, we measured the current draw of 2-prong devices such as our phone and laptop chargers. Therefore, a further complication to accurately measure the AC current draws of our devices would have been to cut open our charging cables, which was out of the question considering they are our important personal devices.
Accomplishments That We Are Proud Of
We are particularly proud of our ability to have found and successfully used sensors to quantify power consumption in our electrical devices. Coming into the competition as a team of mostly strangers, we cycled through different ideas ahead of the Makeathon that we would like to pursue, and 1 of them happened to be how to reduce wasteful power consumption in consumer homes. Finally meeting on the day of, we realized we wanted to pursue the idea, but unfortunately had none of the necessary equipment, such as AC current sensors, available. With some resourcefulness and quick-calling to stores in Toronto, we were luckily able to find components at the local electronics stores, such as Creatron and the Home Hardware, to find the components we needed to make the project we wanted.
In a short period of time, we were able to leverage the use of MongoDB to create an HMI for the user, and also read values from the microcontroller into the database and trend the values.
In addition, we were proud of our research into understanding the operation of the AC current sensor modules and then applying the theory behind AC to DC current and voltage conversion to approximate sensor readings to calculate apparent power generation. In theory the physics are very straightforward, however in practice, troubleshooting and accounting for noise and error in the sensor readings can be confusing!
What's Next for SmartBar
We would build a more precise and accurate analytics system with an extended and extensible user interface for practical everyday use. This could include real-time cost projections for user billing cycles and power use on top of raw consumption data. As well, this also includes developing our system with more accurate and higher resolution sensors to ensure our readings are as accurate as possible. This would include extended research and development using more sophisticated testing equipment such as power supplies and oscilloscopes to accurately measure and record AC current draw. Not to mention, developing a standardized suite of sensors to offer consumers, to account for different types of appliances that require different size sensors, ranging from washing machines and dryers, to ovens and kettles and other smaller electronic or kitchen devices. Furthermore, we would use additional testing to characterize maximum and minimum thresholds for different types of devices, or more simply stated recording when the devices were actually being useful as opposed to idle, to prompt the user with recommendations for when their devices could be automatically shut off to save power. That would make the device truly customizable for different consumer needs, for different devices.