Skip to content

Predicting Forces acting on drones using sensor data

Notifications You must be signed in to change notification settings

JLin37/Environment_Sensing

Repository files navigation

Environment_Sensing

Our research into the use of sensor data from drones is based on Dr. Gary M. Weiss’s work in the Fordham University’s WISDM (Wireless Sensor Data Mining) Lab is concerned with collecting the sensor data from smart phones and other modern mobile devices (e.g., tablet computers, music players, etc.) and mining this sensor data for useful knowledge. The Lab’s work on activity recognition allows them to recognize many of the physical activities that a smartphone user is performing (walking, jogging, sitting, etc.) based on the user's movements, as measured by the smart phone's tri-axial accelerometer.

Our approach is similarly using the accelerometer and gyrometer data to predict the direction and speed of wind acting on a drone in real-time. We explore the use of hand-crafted feature along with the use of CNN to try and find the best model to make such predictions.

Dr. Weiss’s research publication links can be found here: http://www.cis.fordham.edu/wisdm/includes/files/sensorKDD-2010.pdf http://www.cis.fordham.edu/wisdm/includes/files/btas10.pdf