1 - Motivation


A decision-support tool for architecture

Software Problematics

A decision-support tool for architecture

Tools

A decision-support tool for architecture

Description

A decision-support tool for architecture

Example : Mars500/Sirius-17 experiment

A decision-support tool for architecture

Prototype

Motivation

Proposition

Design constraints

Design constraints

New knowledge for built environments

Human aspect integration

Conclusions

Thank you!

Development of a decision-support computer tool 

driven by Artificial Intelligence processing 

to find optimal solutions 

of non-intuitive architectural issues

2 - Proposition

3 - Design constraints (1)

4 - Design constraints (2)

5 - Software problematics

6 - Tools

7 - Description

8 - Example - Sirius-17

9 - Prototype

10 - Perspectives

Thank you !

11 - Conclusions

 

 

 

 

 

 

 

 

 

 

 

Game engine softwares like Unreal Engine or Unity, currently used in the video game industry, are also poweful programming platforms providing a wide set of sophisticated tools :
             > for 3D architectural modelisation,
             > for Artificial Intelligence programming,
             > for character animation and behavioral simulation,
             > for physical simulation (gravity, mecanics, collisions),
             > and compatible with high-level programming langages, such as C++ for Unreal Engine 4 (UE4), or C# for Unity.

For these reasons, we decided to use UE4 as a software platform for our research
3°/ An game engine (Unreal Engine 4) driven by Artificial Intelligence can simulate:
- elementary behavioral modeling of characters (Behavior Trees), 
- interaction with the environment (Environment Query System/EQS), 
- physical constraints (Physical Simulations)

   This prototype includes key concepts which can be developed in future works:

        > C++ implementation of the genetic algorithm in the game engine


        > Enrichment of the 3D library with new modular elements 


        > Addition of new constraints in the Fitness Function


        > Improvement of the A.I. thanks to neural network, 

           with a framework such as TensorFlow 

Human Predictive Simulation
for Earth and Space Exploration

Tatiana Volkova, SSC, EPFL
& Dr Olivier Boisard, OB-Conseil




 

Perspectives

A decision-support tool for Architecture

                              Project
cost constraints



 Challenges :

Technology complexity

Design of complex interdisciplinary projects located in extreme environments 


 


Development of computer decision-support & design assistance tool 

for extreme environments simulations

Accessibility to remote places in extreme environments  

 

 

 

 

 

 

  


 

 

WHAT do we present?

                               


Extreme environments physics simulation

Design decisions optimization

Reducing conceptual design cost



Sensory-related functions



Safety-related functions



Service-related functions




Enhancement-related functions










These projects provide experimental data on the health and performance of astronauts in the conditions of prolonged isolation

 HOW

 to quantify human factors in the software ?

 

Concordia
                                                         ESA 
                                                          Moon Village
                                                                   Tektite I

 

 

new content

2°/  A Fitness Function quantifies and optimizes one - or several - purpose-related functions (security, optimisation of interactions, economical constraints, etc...)
4°/ Fitness Functions are used in genetic algorithms to guide simulations towards optimal design solution
1°/ An architecture based on modular design approach

 

 

 

 

 

 

 

 

 

 

 

   

   

 

A tool with 4 components :

Functional purpose


Values & priority measures

Purpose-related functions

Object-related processes 

Purpose-related functions

Physical objects

.

.

.

.

.

.

.

.

.

The proposed architecture decision-support tool:

> Decreases the number of architectural concepts to be explored


> Enables consistent configuration optimization: conceptions comparison


> Provides a human-centered design


> Speeds up the design process


> Helps to run the different scenarious of future conception usage

tatiana.volkova@epfl.ch

olivier@olivier-boisard.net

Thank you for your attention!

 

.

The gaps identification in the research of past&present 
Earth/Space habitat analogs will lead us to the improved future ones

.

-

-

new headline

new content

.




Period: 2007-2011

Objective: Mars mission simulation

Participants: Russia (IBMP)*, ESA, China 


-




Period: 2017-today
Objectives: 
> 2017- 17days Moon mission 
> 2018 - 4 months mission 
> 2019 - 8 months mission
> 2020 -12 months mission 
Participants: Russia (IBMP)*, 
NASA




Period: 2007-2011
Objective: 105/520-days Mars mission simulation
Participants: Russia (IBMP)*, ESA, China 




SIRIUS logo Credits:IBMP

IBMP - Moscow Institute of Bio-Medical Problems

new headline

new content

new headline

new content

.

.

.

Period: 2007-2011
Objectives: 
> 2007 - 14 days Mats mission
> 2009 -105 days Mars mission
> 2011 - 520 Mars mission
Participants: Russia (IBMP)*, 
ESA, China

WHY Mars500/Sirius?

Moon/Mars Lander Module (EU-50)

Medical Module (EU-100)

Habitable Module (EU-150)

Utility Module (EU-250)

 Accessible experimental technical/human factor data results
International project
Currently in use

 

 

Elon Musk Mars settlement 
Mrs. Tatiana Volkova
Co-Authors