Space Apps Challenge 2017

Earth has been dependent on solar energy from the very beginning. But we, as humans, have only begun to harness that same energy source. As our use of renewable energy gains momentum, it is also our responsibility to manage our energy efficiently and move towards a greener tomorrow. For our project, we focused on solar energy at the HI-SEAS habitat. The day-to-day lives of the crew at HI-SEAS rely on solar energy, and our goal is to make that energy easier to control and use. We created STARLYTE: an artificial intelligence unit integrated with a simplified user interface to allow for increased control of energy consumption and energy efficiency.

Interacting with STARLYTE is simple. Using voice commands, you can easily obtain information from STARLYTE. STARLYTE will then respond to your request with both in audio as well as visual displays.

STARLYTE utilizes real-time weather data and forecasts to predict short-term energy production expectations and willl guide users in energy consumption and conservation. During times of expected decreases in solar energy output, STARLYTE can ensure energy efficiency is maximized by recommending changes in energy use. A map of the HI-SEAS habitat with a color-based data overlay displays current energy use and allows users to easily visualize energy consumption. If solar batteries reach critical levels, energy can be limited to essential units only with a manual override for any non-essential units.

STARLYTE will be continuously monitoring energy output, and can relay information about the state of the solar panels given expected and true output data. This will simplify the maintenance of the solar panel area by alerting HI-SEAS members when the solar panels are malfunctioning. STARLYTE MOBILE, a unit integrated into the space suit helmets can further simplify maintenance by quickly identifying the specific panel that requires repair, and giving real-time feedback about the energy output to ensure the repair is complete.

The deep learning aspect of STARLYTE allows the technology to adapt to different users and environments.

STARLYTE's simplified user interface will allow for expanded applications of our technology outside the HI-SEAS habitat. Homes, schools, cruise ships, anywhere that can use solar energy can benefit from STARLYTE!

Additionaly, STARLYTE offers an increased understanding of and control over energy consumption and can ultimately increase our world's overall energy efficiency. To show real solar panel data, we created a solar panel demonstration using a SunnyTech solar panel and 60W lamp paired with a display created in LabVIEW to show the amount of energy that can be generated with a solar panel and relate that to how much energy would be generated by a solar panel array located at the HI-SEAS location. We expanded on this with another LabVIEW program combined with a python script that compares real-time weather conditions between multiple locations. We compared our location (Fayetteville, NC) with the HI-SEAS location and outer space to find the amount of energy that we could obtain from the same solar array in all three locations.

Our open-source prototype of STARLYTE is fully functional and can be downloaded on our GitHub page listed below. In the future, we would like to gather user input to streamline what data is displayed, develop STARLYTE's learning capabilities to understand patterns of energy use, and to add data sharing to compare energy use between different locations on earth.

Please visit our "Updates" page to see pictures of everything in action!

Our open-source code and diagrams are available here:
https://github.com/MakeItHackin/STARLYTE

Resources Used:

• Special adapted version of “You are my sunshine” by Ben Therrell.

Fully Functional Starlyte Artificial Intelligence Energy System:

• National Solar Radiation Database (NSRDB) – data used from April 30, 2015 for HI-SEAs location, provided by the National Renewable Energy Laboratory
• Meteorological Data from the HI-SEAS Habitat
• Additional solar energy research by PVWatts website
• LabVIEW
• Microsoft Speech Recognition Engine
• Microsoft Speech Synthesizer Engine

Starlyte Mobile:

• RainMeter Desktop App
• Iron Man, Tech PC, Neon Space rainmeter skins

Solar Simulation Demo:

• 60W Swing Arm Lamp
• Realistic fine Mars regolith
• Fog machine
• 1/8” plywood, with custom laser cutting and etching
• 20 – 5mm LEDs (pink and blue)
• Arduino Uno clone
• Relay shield
• SunnyTech Solar Panel, 5 V , 250mW
• GY-521 / MPU6050 – Six Degree of Freedom sensor (3 axis gyro, 3 axis accelerometer)
• Adafruit INA219 High Side DC current Sensor
• Adafruit MicroUSB lipo battery charger
• Adafruit 500 mAh lipo battery
• Switches
• Wire
• 51 Ohm resistors

Weather Location Comparison Demo:

• LabVIEW
• Python
• Wunderground API
• Comparing HI-SEAS, Fayetteville and Space

Simulated Starlyte GUI for Video:

• RainMeter
• Iron Man, Tech PC, Neon Space rainmeter skins
• Projector/LED TV