aquaris

Aquaris is a game ded­i­cated to impart knowl­edge about renew­able energy tech­nolo­gies and energy-saving mea­sures to chil­dren of the fourth grade. It was intended to be set up as an instal­la­tion in a small museum, owned by an elec­tric­ity pro­duc­ing com­pany. Pri­vate guided tours about the topic men­tioned before where offered for whole fourth grade classes at once and should be extended by some more fun part in the end, where the chil­dren can employ the new knowl­edge, moti­vat­ing them to pay atten­tion at the tour before.

Click to see the project pre­sen­ta­tion video (52mb) — also avail­able in lower qual­ity (17mb).

Down­load a key­board con­trolled demo ver­sion (extract data from ZIP, run included aquaris.swf, Flash Player required).

my respon­si­bil­i­ties and tasks

Pro­fes­sor Frog, the host

• lead­ing soft­ware development
• devel­op­ing server soft­ware (see below), par­tially Flash interface
• carry out research about opti­mal con­troller devices to play the game
• cre­at­ing host character
• coor­di­nat­ing graph­ics — audio — code interoperability
• assist con­struc­tion of con­troller devices

require­ments

From the sit­u­a­tion described in the intro­duc­tion and the fact, that by default, a whole class of up to 30 chil­dren takes a tour at once and finally approaches the game, where the pupils should not spend more than one hour, three require­ments were derived:

easy to learn game flow

• sounds and col­ors clear­ify boni and mali
• min­i­mal com­plex­ity: col­lect water and boni for point, avoid mali

easy to use con­troller, which demand exercise

• form fol­lows func­tion: the cur­sor on the screen should fol­low the move­ment of the player
• no but­tons should be nec­es­sary to sim­plify usage
• the player’s posi­tion should rep­re­sent the cursor’s posi­tion on screen to alle­vi­ate the iden­ti­fi­ca­tion of the cur­sor controlled
• the main menu to start the game will be con­trolled by an addi­tional mouse, which allows the guide to super­vise and interfere

impart knowl­edge within 12 minutes

game mode screenshot

• The game was design to be ful­filled in a pre­de­fined time of max­i­mal 12 min­utes and can be played simul­ta­ne­ously by up to six play­ers divided into three groups, i.e. five groups of six peo­ple fit into one hour of total time. But nev­er­the­less after mind­ful lis­ten­ing through­out the whole tour, at least six of the 12 min­utes should be fun only — called game mode.
• 

  quiz mode screenshot

  The idea behind the game is to moti­vate the pupils to pay atten­tion to the tour guide, there­fore knowl­edge pays off in the quiz part. All points col­lected in each part weight 50%, for this rea­son it is impos­si­ble to win by only play­ing well with­out knowl­edge from the tour and vice versa.

• To cope with minor read­ing prob­lems of fourth grade pupils, text and instruc­tions are mostly spoken.

con­sid­ered con­troller designs

Con­di­tions, mate­r­ial, costs and a result­ing con­clu­sion have been gath­ered for each idea.

pressure-sensitive mat

• ZTiles is a mod­u­lar sys­tem of pres­sure sen­si­tive tiles made of rub­ber; weight is mea­sured and trans­mit­ted as ana­log sig­nal; tiles can be clipped together and get con­nected to each other automatically
• could not be bought at that time as still under development

switch­ing mat

• mat con­tains two layer, which get con­nected under pres­sure result­ing only in a on/off dig­i­tal signal;
• very impre­cise track­ing due to dig­i­tal sig­nal sig­nal out­put, not wear-proof — there­fore declined

cam­era tracking

• cam­era needs to be set up, floor prepa­ra­tions costly (needs dull fin­ish), com­plex and error prone soft­ware nec­es­sary, play­ers need to wear col­ored clothes to dis­tin­guish them, install cur­tains to cre­ate a con­stant light situation
• in con­clu­sion, error prone player iden­ti­fi­ca­tion and the  sys­tems is vul­ner­a­ble to any kind of dirt
• costs are rel­a­tively low; approx­i­mately 400€ for the cam­era, 1000€ to pre­pare the floor, some cloth­ing; 1500€ in total

han­dle with cable pull mechanism

• com­plex cus­tom con­struc­tion, poten­tiome­ter could mea­sure posi­tion of han­dle by count­ing rota­tions of a guide roll; com­puter hard­ware needed to con­vert cur­rent flow into dig­i­tal signals
• very pre­cise track­ing, play­ers are forced to take exercise
• com­puter hard­ware about 200–500€, one con­troller per player about 500€; 3500€ in total

see­saw

• quite sim­i­lar tech­ni­cal require­ment to the pre­vi­ous device, instead of a rope with han­dle pulled by the player, he steps upon a see­saw and the angle is measured
• an exper­i­men­tal setup showed, that it is very dif­fi­cult to reach a spe­cific posi­tion and hold it
• the con­struc­tion should be sim­pler and there­fore cheaper, around 300€, result­ing in 2300€

RF-ID chip tracking

• sen­sors need to be placed around the play­ing area, at least three or four, but six should be much more accu­rate; as the cov­er­age of pas­sive chips are too low (below one meter), semi-active ones, which auto­mat­i­cally switch to standby, to save energy, which allows them to work for more than five year with one bat­tery set, would be needed (six meters of range and more),
• each sen­sor is able to iden­tify eighty steps of sig­nal strength, by using at least three sen­sors, the posi­tion of each chip should be cal­cu­la­ble; as the sig­nal strength is highly depen­dent on envi­ron­ment changes (play­ers occlude each other), an amount of six sen­sors to cover the whole field would be necessary;
• addi­tion­ally the player needs to “wear” the chip some­how, maybe hid­den in a cap or in shoes
• one sen­sor costs about 1200€, each chip about 30€; 7400€ in total.

deci­sion

RF-ID track­ing was very appeal­ing, despite the fact that we would have needed to exper­i­ment a lot, though quite expen­sive. But as our spon­sor left the project for inter­nal rea­sons, we had to cut down on bud­get. Luck­ily we could win over a mecha­tron­ics stu­dent and go for another very promis­ing idea, the han­dle with cable pull device, as we con­sid­ered it to be the most con­ve­nient alter­na­tive. The han­dling is very close to the move­ment of the cur­sor on the screen and we regarded it to be the most quickly and eas­ily learnable.

advan­tages at a glance

model of the con­troller device

• form fol­lows func­tion: the cur­sor on the screen fol­lows one to one the con­troller handle
• no but­ton func­tions need to be learned, all inter­ac­tions nec­es­sary to play the game can be per­formed mov­ing the handle
• each con­troller device is placed accord­ingly to its cur­sor on the screen
• very pre­cise
• needs to be cal­i­brated only once at setup time
• no influ­ences of environment

See a 3D ani­ma­tion (4mb) about place­ment and func­tion of the con­trollers or its low qual­ity ver­sion (1mb).

real­iza­tion

The new con­troller in a very early state on the right side.

First con­troller pro­to­type with a width of approx­i­mately two meters

Con­troller con­nected to the inter­nal A/D con­verter board.

See a video about set­ting up the con­trollers (20mb wmv) for the pub­lic project pre­sen­ta­tion — low qual­ity (5mb wmv).

hard­ware

soft­ware

esti­mated total costs

Finally we cal­cu­lated (in 2006) the costs to get all con­troller devices pro­duced by a com­pany plus addi­tion­ally nec­es­sary equipment.

Though we could not effort the con­struc­tion of a full set of six con­trollers, two pro­to­types have been built, to demon­strate the functionality.