SolarX

• 

  Our assembled chassis with all components.

Inspiration

With commercial solar panels operating at efficiencies between 17-20%, maximizing the light exposure that these panels get plays a critical role in the global transition to sustainable energy.

Driven by our interests in electronics and sustainable energy sources, and wanting to get exposure to the field of controls, we took on the UTRA Hacks Open Challenge to forge our problem statement concerning the automation and optimization of light exposure to solar energy, allowing us to uniquely build a robot reflecting our interests in solar power conversion, microcontroller programming, and sensor data processing.

What it does

The SolarX rover is controlled by an Arduino Uno that reads photoresistor inputs from the front and sides of the chassis. These inputs feed into a control loop that calculates the rate at which each 12V motor is spun, based on a PWM signal from the L298N driver. The goal of this prototype is to have the rover make positional adjustments to expose the housed solar panel to more light. In the demo, we have the robot chase a flashlight, eventually settling in once the light is exposed over the panel! There is a flashing LED on the back of the bot, that indicates to us the panel operation state, and how well it is converting the incoming light.

How we built it

We started with the UTRA Open Challenge kit, which sourced the Arduino Uno, 9V battery, photoresistors, motor controller, and breadboard. We then sourced a voltage regulation board, allowing us to externally supply 3.3V and 5V, powering the Arduino without a USB connection. We also used a small solar panel that is roughly the size of the SolarX rover, supplying anywhere from 1.5V to 4.5V. This solar panel became the focus of the project. Finally, the main structure of our rover ended up being an aluminum and polycarb chassis, with treads powered by a 12V motor, one per side.

Challenges we ran into

Without a doubt, one of the greatest challenges we faced was calibration and tuning of the robot's desire to chase light. We found that our photoresistors, the main drivers for the robot's perception of light, had wildly varying levels of sensitivity between them. This was one reason we found it difficult to correctly and consistently control the robot's behavior. Another challenge was our lack of knowledge in the area of control systems. While PIDs (proportional-integral-derivative controllers) were something we investigated as a potential driver for our robot's behavior, we feel our implementation of one was rudimentary and bounded by our knowledge on the topic.

Accomplishments that we're proud of

As we dove into this problem statement, we were excited to tackle something linked closely to our interests but were worried about our chances of producing a successful product. Several concerns arose throughout the process, like how to mount the solar panels given a lack of mechanical components (i.e. chassis, wheels, etc) and also the integration of the panel into the design. However, we are proud of how we navigated each of these hardships. We believe the resulting product of this hackathon speaks to our team's ability to overcome obstacles in a cooperative and efficient way.

Additionally, we feel that a great deal of technical expertise was gained in creating this project. Interfacing with solar panels was not something many members of this team had experience with, and it was very educational to see firsthand how one would work with them.

What we learned

SolarX gave us experience researching and designing with a variety of sensors, including the collection of ambient temperature, pressure, and altitude over I2C, as well as IR digital readings, and basic stepper and servo motor testing. Furthermore, we explored a variety of power conversion methods, from the receiving solar panel energy to multiple LDOs, having a distributed power architecture was a fun system to design for the integration of our various hardware. Finally, we began to explore the broad and intensive field of controls, where we played around with tuning our feedback loop that focused on modifying the motor PWM based on the derivative and present readings being found on our photoresistors. The next step would be to properly implement a proportional value, along with the integral.

What's next for SolarX

A self-orienting system will need a high angular resolution for the best alignment to maximize the solar panel's exposure to the light source. More photoresistors distributed around the vehicle will be needed to capture this optimal position. In addition, control over the solar panel’s angle with respect to the horizontal plane can achieve alignments with more panel exposure depending on the sun’s position in the sky.

Built With

• arduino
• github