Documents

Team Presentations & Reports

NamePresented onPresentationReport
Project Progress on Systems EngineeringOct 29, 2024Link
Conceptual Design ReviewDec 5, 2024LinkLink
Progress Review #1Feb 13, 2025Link
Progress Review #2Feb 27, 2025Link
Preliminary Design ReviewMar 11, 2025Link
Progress Review #3Mar 20, 2025Link
Progress Review #4Apr 8, 2025Link
Critical Design ReviewApr 28, 2025Link

PCB Assignment

Since our drone platform is built from the Hexsoon EDU-650 frame kit, a custom power distribution board (PDB) was included with the system. Instead of replacing the professionally-made PDB, we decided to implement another function that is critical to our project: battery voltage monitoring.

Ensuring that a LiPo battery is within its operating voltage range is a safety-critical application. The safety requirement of the operating voltage for a LiPo battery is between 3.2V/cell and 4.2V/cell. The ideal operating voltage per cell for a LiPo battery is around its nominal voltage of 3.8V/cell. Overvolting (some cell has a voltage greater than 4.2) and undervolting (some cell has a voltage less than 3.2V) cause permanent damage to the battery, reduce its lifespan and make it susceptible to volatile behavior such as smoking or catching fire. However, overvolting a battery can only happen during charging and we rely on good quality chargers that balance the battery while charging. Hence, we will focus on preventing undervolting.

Our team is planning on making a printed circuit board (PCB) that can do the following:

  • Detect the voltage of each cell in a 6 cell (6S battery)
  • Provide a visual and an audible alarm when any cell reaches below 3.2V
  • Have an LED display that shows the current voltage of each cell and total battery voltage in rotation.

Schematic

Layout

Individual Learning Reports (ILRs)

Aayush FadiaAbhishek IyerAmy JiangRanai SrivastavSwastik Mahapatra
ILR01ILR01ILR01ILR01ILR01
ILR02ILR02ILR02ILR02ILR02
ILR03ILR03ILR03ILR03ILR03
ILR04ILR04ILR04ILR04ILR04
ILR05ILR05ILR05ILR05ILR05

Design Brainstorming

Drawings, schematics, datasheets

Electrical

All electrical documentation, wiring diagrams, schematics, datasheets can be found at this documentation link: https://airlab.slite.page/p/ppKdwPyX3i36hc/Electrical-Connections

Mechanical

Final drone CAD

Datasheets

NVIDIA Jetson AGX Orin Developer Kit Carrier Board Specification

TELEDYNE FLIR BOSON AND BOSON+ THERMAL IMAGING CORE PRODUCT DATA SHEET

Epson M-G365PDC1/PDF1 Data Sheet

Frsky X7 X7S ACCESS Transmitter Manual

Frsky X8R Receiver Manual

Component testing & experiment results

The odometry subsystem was tested on various paths and evaluated by simultaneously collecting ground truth data using the Intel Realsense in a non-smoked environment.
The mapping subsystem was evaluated against a ground truth global point cloud collected by the Intel Realsense in a non-smoked environment.

Software

The codebase is built in ROS2 and hosted in the following Github repos:

  • https://github.com/castacks/smores_drone_software: Houses higher-level algorithms and system-specific configuration files.
  • https://github.com/castacks/smores_drone_drivers: Houses low-level sensor initialization software.
  • https://github.com/MAC-VO/MAC-VO-ROS2: Houses MAC-VO, the onboard odometry algorithm.

The entire software stack can be initialized with the launch.sh bash script in the smores_drone_software repo. This script will open a tmux session that brings up all sensors (thermal cameras, realsense, IMU), initializes MAC-VO odometry with the appropriate configuration, initializes dense depth estimation with FoundationStereo, initializes global mapping via Octomap, and finally opens Rviz and the combined camera feed visualizer. Additionally, the tmux session will include several windows that run ROS debugger commands to the relevant topics such as ros2 topic echo and ros2 topic hz so that the user can verify all components of the software stack are running as expected and easily debug as necessary.