Maria visit's to JPL this Summer

Prof Maria Fasli, co-investigator in the project, visited Adrian Stoica at JPL between the end of July and early August this Summer. Maria is an expert in multi-agent systems and she is looking at mixed-initiative systems for robot control with Ardian's group.

Collaborative Brain Computer Interface for Simulated Spacecraft Control

While at NASA JPL in the Summer 2012 working with Adrian Stoica, Riccardo, Caterina and Francisco experimented with the real-time control of a simulated spacecraft using the brain signals from two users.

Here is the setup

The subject on the left is Riccardo. The subject on the right is Luis Clark, a student volunteer.

Here is a close-up of the simulator

The objective was to pass as close as possible to the planet (intermittently visible in white). Here is a movie of one of the simulations

The development and experimentation of the software had started some months before at Essex and are still ongoing. Preliminary results on this line of research will appear in the Intelligent User Interfaces conference in March 2013 in a paper entitled "Towards Cooperative Brain-Computer Interfaces for Space Navigation". 

Here is the abstract:

We explored the possibility of controlling a spacecraft simulator using an analogue Brain-Computer Interface (BCI) for  2-D pointer control. This is a difficult task, for which  no previous attempt has been reported in the literature. Our system  relies on an active display which produces event-related potentials (ERPs)  in the user's brain. These are analysed in real-time to produce  control vectors for the user interface.  In tests, users of the simulator were told to pass as close as  possible to the Sun.   Performance was very promising, on average users managing to satisfy the simulation success criterion in 67.5% of the runs.  Furthermore, to study the  potential of a collaborative approach to spacecraft navigation, we  developed  BCIs   where the system is controlled via the integration of the ERPs of two users. Performance   analysis indicates that collaborative BCIs  produce trajectories  that are statistically significantly superior to those obtained by single users.