[sci.virtual-worlds] IBM Veridical Interface: Report from CHI

cyberoid@milton.u.washington.edu (Robert Jacobson) (05/31/91)

|IBM: ADVANCE/IBM researchers are using "virtual reality" to explore ways to
|    make computers friendlier
|
|April 30, 1991
|
|    (ADVANCE) NEW ORLEANS--(BUSINESS WIRE)--IBM researchers are giving 
|a subtly spectacular demonstration in New Orleans which suggests how 
|computer users may one day virtually "enter the realm" of their work, 
|whether it be a molecule -- to explore its strucuture -- or a financial 
|record, to study a corporate or national economy.
|
|    The IBM demonstration is a kind of "deep metaphor" of a world 
|beyond the looking glass.  In the demonstration, people in the real 
|world can interact with each other through three-dimentional, moving 
|objects in an artificial world depicted on their computer screens.
|
|    A team of scientists led by Daniel T. Ling of IBM's Thomas J. 
|Watson Research Center in Yorktown Heights, N.Y., is using the 
|demonstration to explore ways in which computer-generated virtual 
|worlds may extend human cognitive and perceptual faculties into the 
|computer and eventually erase the boundary between person and machine.
|
|    Far more than fun-and-games that so-called "virtual reality" -
|simulations of the real world - often seems to be to the general 
|public, a primary objective of the IBM research effort is to make 
|computers easier to use by enabling them to interact more fluidly and 
|naturally with their users.
|
|    On the surface, it is all just a game played in a wire cage shown 
|on each player's screen.  The human opponents - designated "alpha" 
|and "beta" - stand and view the playing area from opposite ends of 
|the cage that tilts differently for each according to the player's 
|perspective, or head position.
|
|    Brightly colored, flexible geometric objects - "rubber rocks" of 
|different shapes - appear spontaneously and bounce around the cage, 
|changing hue as they go, from blue to green to red; shortly after they 
|turn red, they "explode." 
|
|    The object of the game is to bat, squirt (with a jet from pointed 
|forefinger) or grab an object and move it near the opponent before it 
|explodes and subtracts a point from that player's score.  
|
|    No sci-fi headgear with goggles and earphones are involved.  
|Three-dimensionality is created by perspective and movements of the 
|players' heads and of the objects on their computer screens. 
|
|    A sensor on the cap each player wears registers head position as 
|"motion parallax", making the cage move.
|
|    Each player wears a glove that signals hand movements and gestures 
|(pointing and grabbing).  The "game", itself, can follow verbal 
|commands, and it announces the score in a synthetic voice ("got alpha; 
|alpha four, beta five"), as well as producing sounds of rocks bouncing 
|and breaking in the cage.
|
|    Everything in between is done by seven IBM RS/6000 Power 
|Workstations connected in a way both to amass computing power and to 
|distribute it so that the two players could as well be five or six 
|(with additional RS/6000s) and located not in the same room but in 
|different places across the country or around the world.
|
|    While the game is fun to play and to watch, it achieves its subtle
|spectacularity by what it represents beneath its flashy exterior; a 
|way, eventually, both of making computers more widely and easily accessible
|to people everywhere and of enabling groups of scientists, engineers, 
|economists or other specialists virtually to "get into their data" 
|and collaborate on a problem from several different remote locations 
|simultaneously.
| 
|    Dr. Ling and his colleagues are presenting their work at a meeting 
|of the Association for Computing Machinery's Special Interest Group in 
|Computers and Human Interaction (SIGCHI) at the Sheraton New Orleans Hotel.
| 
|    The chief novel feature of the IBM simulation is the "dialogue 
|manager".  This system coordinates each player's movements and his or 
|her interactions with the objects, keeping separate what appears on the 
|screen from the mechanisms that produce it, thus separating cause and 
|effect in the simulation of interactions between players and the 
|virtual world.  The movement of the objects is handled by the 
|simulator.
| 
|    The cause-effect separation is demonstrated in the creation of a 
|rubber rock.  Rocks can be created in three ways: automatically during 
|the game, by voice command, or manually by selection from a menu with an 
|input device called a mouse.
| 
|    The dialogue manager "reads" all three mechanisms of ordering an 
|object but tells the simulator simply to make one without having to 
|specify how the order had been given.
| 
|    "This makes it possible to have multiple persons in the same 
|virtual world cause things to happen which are different from what one 
|person could do," said Dr. Ling
| 
|    For example, two players can pick up the same rock and break it 
|apart, thereby interacting with each other through a simulated object.
| 
|      Such feats are produced by a collaboration of IBM RS/6000 Power
|Workstations.  One produces the graphic representation of the playing
|area for each player; that's two.  Another simulates the "rubber 
|rocks" and the way they wobble and bounce and respond to actions of 
|the players in a physically realistic situation in real time.  
| 
|    Two more manage the dialogue for the players, translating their 
|gestures and hand movements and also producing the game's "own" 
|speech (comments, score-keeping).  A sixth RS/6000 handles input from
|the glove and head-tracker and recognizes gestures.  The seventh 
|recognizes spoken commands for creating objects and coloring them.  
| 
|    Dr. Ling explained that this demonstration is the third in a 
|logical progression of improvements in artificial worlds at IBM.  The 
|first project was a simulated handball game involving a glove with 
|which the player could bat a ball around a simulated room: a very 
|simple simulation, he said.  
| 
|    The second project was more complex.  In it, a fluid vortex tube 
|was created mathematically; the simulation involved grasping the tube 
|(data), turning it and even moving through it.  This project brought in 
|spoken commands and more complex gestures.
| 
|    The new "rubber-rock game" capitalizes on experience with the 
|previous models, enabling multiple users to interact with each other 
|and with a complex simulation produced with multiple channels of data 
|(graphics, sound, motion) simultaneously from multiple locations.  The 
|scientists say they will eventually incorporate touch, and, perhaps, 
|even smell.
| 
|    Associated with Dr. Ling in this work at IBM were: Christopher F. 
|Codella, Ronald I. Frank, Reza Jalili, Lawrence Koved, Bryan Lewis, 
|Alan Norton, David Rabenhorst, Paula K. Sweeny, G. Turk Turk (now at 
|the University of North Carolina) and C.P. Wang.


--