The aim is to develop a tool to monitor and detect elements of the crew psychological state and social well-being through voice and linguistic content analysis of open channel communication, without interference to everyday operations.

 

 

Long-duration exploration missions like e.g. to Mars will pose completely new challenges to space crews, compared to Low Earth Orbit flights. One of the potentially critical issues is psychology. Multiple stressors act on the minds in spaceflight, based on the general environment, mission characteristics, the habitat, the social situation etc. All of these will be present or even much more prominent in exploration class missions. Most psychological support measures in use today are employed in-flight. Examples are regular private conferences with psychologists, regular conferences with family, resupplies, uplink of news, visiting crews etc. However due to communication delays, probable bandwidth restrictions and mission characteristics (that e.g. don?t allow for resupply or visiting crews) most of these measures will either be impossible, or possible only in a limited measure.

 

Research work has demonstrated that changes in physiological state, emotional state, cognitive load and fatigue can be detected through their effect on physical measurements of the body.

 

This proposal aims at covering the phase one of a three-phase development activity dealing with astronauts psychological state assessment. The 3 phases are namely:

 

• Preparation for the technological toolset required to support a joint (Russia-UK) ISS experiment, including test and advancement of the state-of the-art techniques and technologies in voice and content analysis using MARS500 data.

 

• Technical support of a 18-month ISS experiment (2015) including test of state- of-the art techniques through collection and analysis of live open channelcommunication data for voice and content analysis of crew communication intra- and inter-team to improve the state of the art.

 

• Software development such that an operational system can be implementedon ISS.

 

This activity will cover the phase 1 and the proposed work is the following:

 

The voice analysis method will be based on tracking the voice characteristics of each individual crew member / subject over a long period to build up a statistical model of the natural variation in their speech. These voice characteristics will be derived from audio recordings of their speech made in everyday interactions, and will include parameters such as speaking rate, intensity, voice pitch and pitch variation, irregularity, breathiness, spectral slope and spectral modulations. This will be obtained by preparing a MARS500 voice and linguistic content corpus.

 

An existing state-of-the art voice analysis toolbox will be used and enhanced to derive voice parameters from the audio, and statistical methods for the analysis of time series will be used to detect voice events - being changes to the expected voice statistics occurring on particular occasions. One challenge will be theadaptation of the technology to make it work with English content (currently working for Russian language).

 

After testing, the upgraded voice analysis toolbox will then be used for validation purposes by analysing over 3000 video messages sent by crew members on the Mars500 mission. The validation will seek to demonstrate how robust voice parameters can be extracted from the audio channel on these recordings and how events (such as the simulated landing on Mars) can be detected automatically from changes in the voice statistics for the crew.

 

The outcomes of the proposed activity are expected to be useful both for space research and terrestrial applications (monitoring of stress and fatigue for safety- critical jobs, such as air traffic controllers, truck drivers in large open mines, etc).

 

The activity will be carried out according to a clear set of conventional tasks:

 

• Task 1: Literature review and background analysis of existing technology.

 

• Task 2: Workshop with key experts in voice analysis for both space and terrestrial relevant environments (e.g. pilots and Antarctic explorers).

 

• Task 3: Long-duration Mission Requirement Consolidation.

 

• Task 4: Technological Concept Design.

 

• Task 5: Development and Validation.

 

• Task 6: Synergies with terrestrial applications.