Week 4: Research Proposals!

 Hello everyone! 

   This week, I focused on volunteering for the Valentine's Day Marketplace. It was a great success, and it was nice seeing some of you there! Thank you to everyone who bought items, you guys rock! Other than that, I finished my research proposal for L-BEAR. Here is an excerpt from our methodology section: 

The development of the L-BEAR will involve several stages, including design, fabrication, and programming. First, a conceptual design will be created, considering cargo space needs, a contraption to activate door and elevator buttons, and a user interface to control the machine. This design will be illustrated using 3D modeling tools to determine measurements. The design will be refined in the model to ensure optimal placement and sizing before the next fabrication stage.

The fabrication of L-BEAR will involve the integration of motors, sensors, and a 3D-printed chassis utilizing PLA+ filament. The motors will be needed for movement, which sensor data will mainly determine. The PLA+ filament was chosen for its rigidity, heat-resistant properties, and ease of fabrication. All components will help L-BEAR assist with weight-bearing activities and transportation. The ASU will be designed to be modular, allowing for easy maintenance and upgrades.

Finally, programming the L-BEAR will involve creating a user-friendly interface that can be controlled using a tablet and will be programmed to navigate campus pathways and detect obstacles, like pedestrians, objects, and our on-campus cats. Work will start with programming a programmable rover which will eventually be integrated into the final iteration, utilizing initial sensor data and recalibrating it. Machine learning integration will utilize Microsoft Azure to interpret sensor data and create a neural network accordingly. Most data collection will be during the programming phase as we hope to create a robot with a 90% accuracy for the detection of common objects. We hope to discover that these technologies can be developed at Phoenix College to add to public and open-source knowledge. 

Variable Table: 

Name	I/D/C	Symbol	Units	Description
Algorithm for Cat detection	Independent	ALGC	-	Algorithm for campus cat detection
Algorithm for General Common Objects	Independent	ALGG	-	Algorithm for common object detection
Weight capacity	Independent	WC	kg	Weight capacity of the robot
Negative Control- no video feed	Control	NC	-	Negative control with no video feed will be used for calibration and accuracy testing purposes
Positive Control- cat photo	Control	PC	-	Positive control of cat photo to determine if algorithm can differentiate 3D objects from images.
Accuracy	Dependent	Acu	%	Accuracy of detection for each algorithm