Implement a drone capable of autonomously navigating through a defined obstacle course using onboard sensors and control algorithms
Obstacle Detection and Avoidance
Integrate and calibrate sensors ranging from LIDAR to optical flow to ensure obstacle detection of varying shape, size, and placement
Reliability and Control
Ensure stable flight dynamics when exposed to environmental disturbances
Develop a control system that balances agility with accuracy to ensure proper movement
Extensibility
We want the system to be designed in a way that allows for future additions, such as a delivery mechanism, the ability to work with OpenCV, or outdoor deployment
Document the hardware and software framework so they can be adapted for future applications
Challenges:
Flight time
One of the most pressing challenges is the drone’s restricted flight time. The added weight of sensors and structural components required for obstacle detection further reduces the endurance of the system. Any miscalculation of the power budget could result in the drone running out of battery mid-course and hardware damage.
Careful power management and efficient design choices are essential to maximise flight duration.
Sensor reliability in complex environments
The obstacle course introduces surfaces, textures, and lighting conditions that may confuse the onboard sensors. Lidar may struggle with transparent or black materials, cameras may falter in low light, and the ultrasonic sensors can bounce signals incorrectly.
If these failures occur mid-flight, the drone’s ability to avoid collisions is compromised. Robust sensor fusion will be required to maintain accuracy.
Control System tuning for maneuvers
Achieving precise control in an obstacle-rich environment is fairly difficult. An overtuned system risks oscillations and instability, while an undertuned one responds too slowly to avoid obstacles.
The challenge lies in finding a smooth middle ground between agility and stability, ensuring the drone can fly smoothly while reacting quickly to changes in the environment.
Integration of subsystems under stress
The drone must integrate mechanical design, power electronics, sensors, and software into one seamless system. Under the stress of obstacle navigation, any mismatch in timing, power draw, or communication can cause cascading failures.
Ensuring subsystem compatibility and resilience under real test conditions will be one of the most difficult engineering hurdles to overcome.
