Drones
classic approach of throwing sensors at the problem
optical flow sensor
An optical flow sensor is a vision-based device that captures sequential images of the ground to calculate relative velocity and movement direction for stabilization and navigation
They function similarly to an optical computer mouse, taking rapid low-resolution images of the surface below and measuring pixel displacement between frames. [1] (https://www.unmannedtechshop.co.uk/blogs/knowledge-base/optical-flow-sensors-drones-autonomous-indoor-flight)
Distance Pairing: Because raw pixel shift only indicates angular movement, the sensor must be paired with a distance rangefinder (such as a ToF or LiDAR module) to calculate actual ground speed and altitude
Sensor Fusion: Flight controllers combine optical flow data with inertial measurement units (IMUs), barometers, and gyroscopes to achieve stable hovering in GNSS-denied or indoor environments.
🤔 the winning team used a simple 2.5D LIDAR and very cheap compute
NASA Ingenuity drone has black and white camera, laser altimeter and on-chip gyroscope
🤔 ❓Payload constraints and power constraints
🤔 ❓study the design decisions
Indian drones used to water plants indoor in GPS denied environments
Other Indian drones used in warehouses Addverb
Custom drones used for vertical lifting
indoor environment
relate it to concepts studies in class
puzzles in beginning of class
🤔 ❓what challenges can come up when you are building semi-autonomous robots doing navigation on Mars?