our plan
In attempts to remedy the worldwide issue of planted landmines, we will develop a low cost, low weight, and agile autonomous vehicle. An emphasis towards designing an adaptable mounting bracket for hand-held metal detectors to allow the operator to place any mine detecting device of their choice, within the robot's constraints, will be enforced throughout our design process.
The cost of our robot will be an essential factor to the success of this project as local humanitarian demining groups generally do not have the required funding and/or resources to attain expensive solutions. Thus, our robot will attempt to utilize as much standard and off-the-shelf items as possible, in order to limit the total cost of the robot.
The total weight of our robot is also a design factor that will be taken into consideration in hopes of creating a robot that can maneuver through mine fields without the possibility of accidental detonation of a landmine. Our robot must not only operate successfully on uneven terrain, but also exert less force per ground contact than is necessary to detonate an anti-personnel landmine (11lbs-35lbs). [5]
The autonomous abilities of the robot will be a behavior-based architecture that is based on layers with different levels of competence. The lowest layer will be in charge of avoiding obstacles, and upper layers contain other task-oriented behaviors. Generally, lower-level actions are given higher priority; to avoid an obstacle is more important for the survival of the robot than to follow a defined path. Through the use of the proper sensors, environmental restrictions will trigger lower-level behaviors that are in charge of avoiding obstacles and maintaining stability. The robot will keep track of any deviations from the predetermined search path brought on by environmental factors. At the conclusion of the mission, the terrain's digital map will clearly indicate areas that were searched and skipped, along with any mine targets that were discovered. [6]
OUR GOALS
- Develop a robust and accurate sensor system for detecting landmines; using metal detection and computer vision. Implement advanced algorithms for data processing and analysis, such as machine learning, to increase the vehicle's detection and classification capabilities. According to studies, the average accuracy rate of landmine detection by deminers using metal detectors is around 80-90%; however, this is an estimated figure due to the fact that this statistic can vary greatly in different types of mine fields where there may be higher rates of metal debris. Thus, we will aim to develop a computer vision system that can detect and accurately classify at least 90% of landmines that are placed above the ground. The landmine that will be tested using 3D printed replicas are the VS-50, PMN, and PFM-1, as they are the most common types of anti-personnel mines used today.
- Create a reliable navigation system that can safely navigate the robot through rough and unpredictable terrain. Simulation and experimentation will be used to test this.
- Develop a user-friendly interface for monitoring and controlling the vehicle, as well as for viewing and analyzing data.