Description

This MSc thesis aims at developing an attitude estimator that will process readings of an inertial measurement unit (IMU) using a Spiking Neural Network (SNN). SNNs were developed in computational neuroscience to replicate the behavior of organic neurons. Similarly, IMUs replicate the vestibular system, which provides the sense of balance and the information about body position that allows rapid compensatory movements in response to both self-induced and externally generated forces. In this thesis, both concepts will be joined, fusing acceleration and angular velocity observations of an IMU using a neuro-morphic estimator based on SNNs to estimate the attitude of the sensor.

The work is under the scope of the following projects:

• EBCON: Motion estimation and control with event cameras (web)
• AUDEL: Autonomous package delivery in urban areas (web)