2026 ASEE Annual Conference & Exposition

Legacy smart residential microgrid systems used for academic purposes often depend on third-party cloud APIs that may eventually become restricted or discontinued over time, causing major system failures and loss of core functionality. This work summarizes two technical and
educational case studies from recent revitalization of __________ University’s smart residential microgrid testbed. Our work illustrates the need for engineering students to learn to apply critical and innovative thinking to plan for and overcome eventual platform obsolescence which too
often quickly affects technological solutions forcing the thoughtful design of resilient local solutions in a rapidly evolving communications and security landscape. In the initial example, a Weather Underground API, originally opted for in 2015 to retrieve temperature, irradiance,
humidity, and pressure data from the application, was retired, halting the system’s ability to automatically retrieve weather-based control information. To achieve similar capability our system needed to be redesigned to use a Davis Vantage Pro2 weather station via a TP-Link
TL-MR3020 router in access-point mode running Meteobridge firmware. This important fix now pushes live weather data every 15 minutes directly to the Raspberry Pi controller of the microgrid. While eliminating cloud dependency it afforded accurate local sensing capability, core
to device real-time algorithm modification when forecasts were different than actual site weather. In the second case, the proprietary Ecobee thermostat API became restricted after the manufacturer suspended new developer registrations for security reasons, preventing further
system communications and integrated control when newer modes of these smart thermostats were installed. In answer to this challenge our team redesigned the communication/control pipeline to pair both the upstairs and downstairs thermostats with the Raspberry Pi through
Apple HomeKit using Homebridge, enabling secure, local communication needed for microgrid prediction and control algorithms. The brains of our microgrid, the Pi, now receives live outside weather data and forecasts as well as internal temperature and mode data from each thermostat
and can send hourly control setpoints to maintain comfort and efficiency for the occupants. These revisions to the system were first operated and tested in simulation mode (DEBUG=True) to verify successful behavior before running in live mode (DEBUG=False) for real-time control.
These experiences highlight the growing risk of API dependence in engineering design and the educational importance of adaptability, creativity, and independent problem-solving on the part of student engineers. Students must be encouraged to anticipate similar (and more regular)
technological discontinuities in our ever more rapidly evolving world, and be trained to consider and build alternative control paths based on what is coming next. Having these experiences can help them more readily transform constraints into opportunities for innovation.