JAE
CSE 788 - SPRING 08


Multi-touch Interactive Fluids Table
(See the blog for current progress)

>> An interactive and collaborative game experience for the exploration of the properties of fluids, smoke, and particles by early childhood learners
>> The objective of the game would be to collaboratively direct a "pinball" using a stream of flow to a particular target using digital manipulatives and hand gestures on a multi touch surface.

fluid pinbal concept image

>> In the process, users would learn about the properties of different fluids and the way that they interact with static and dynamic objects in their environment

    • viscosity
    • friction
    • relative mass & density
    • collisions
    • flow & vectors
    • force
>> Some various flows to explore could be
    • steam
    • water
    • oil
    • lava
    • sand
Current State of the Art

>> Multi-Touch Sensing through Frustrated Total Internal Reflection
     (Jeff Han - SIGGRAPH 05)
>> It seems like SPH would be an ideal framework to use since it can be extended to may different types of flow:
   > Granular material -- Particle-based simulation of granular materials
      (Bell, Yu, & Muhca - SCA 05)
   > Liquid -- Weakly compressible SPH for free surface flows
      (Becker & Teschner - SIGGRAPH 07)
   > Gas - Depicting fire & other gaseous phenomena using diffusion processes
      (Stam & Fiume - SIGGRAPH 95)
   > And there are a wealth of other papers on the subject...
Project Motivation

>> This project would obviously lead to a great demo piece (for ACCAD or CSE) that could also merit a Poster at SIGGRAPH
>> Fulfills a large number of ODE Early Learning Content Standards
>> Could possibly serve as an exploratory installation at COSI or other museum
>> A parameterized SPH flow model could result in a paper
Challenges


>> Real-Time Interactive Rates
>> Multi-Touch Interface Integration>> Stable parameterizable model of SPH

Advantages: This project idea has very reasonable scope, but also provides a fun and educational application with a reasonable amount of work. It should only take 3-4 people and it should be doable by the end of the quarter. There is a wide array of literature and code out there on the topic of SPH applied to many different types of situations. Sundar has a game programming book that provides a ready-made framework for implementing SPH.

Disadvantage: SPH is a field where a lot of work has been done and if a parameterizable model is not possible, it would be hard to publish a graphics paper on this project -- though a poster is likely to come from it, as well as a papers in either something like HCI or an education and technology conference.

Quarter Goals

>> FIrst step - get basic particle system working on Multi-touch table
>> Implement a flexible SPH model
>> Implement collision detection & handling
>> Interface Tweaking