Slide 0-18 Evaluating the Length of Virtual Horizontal Bar Chart Columns Augmented with Wrench and Sound Feedback Tatiana G. Evreinova, Grigori Evreinov, and Roope Raisamo TAUCHI Computer-Human Interaction Unit Department of Computer Sciences FIN-33014 University of Tampere, Finland {e_tg, grse, rr}@cs.uta.fi Slide 1-18 p-2-p communication harness' force feedback channel samples: picture 1: harness racing picture 2: baby harness picture 3: horse bridle picture 4: guide dog harness Slide 2-18 Spider concept Illustrations of the cable-suspended haptic interfaces http://sklab-www.pi.titech.ac.jp/~somsak/spidar8.html http://sklab-www.pi.titech.ac.jp/detail/spidar/recent-e.html Robert L. Williams II Cable-Suspended Haptic Interface. Int. J. of VR, 3(3), 1998, 13-21 Slide 3-18 Pen concept: stick/stylus mechanical feedback to user when cursor passes over icons or whatsoever appears on screen can be added to pen manipulations... picture 1: Computer Pen Interface http://www.katzdesign.com/HTML/HPT.htm Pen Cat pro http://www.cbdp.qc.ca/pencatpro_eng.html Pietro Buttolo Characterization of Human Pen Grasp With Haptic Displays, PhD, 1996 picture 2: Haptic pen - a tactile haptic stylus http://www.merl.com/projects/hapticstylus/ picture 3: Sharmin, S., Evreinov, G., Raisamo R. Non-Visual Feedback Cues for Pen Computing. in Proc. of WHC 2005, Pisa, Italy, 2005. IEEE Inc., pp. 625-628. Slide 4-18 another pen concepts: stick/stylus based on touchscreen, Pulley + DC motor + sensor of thread's tension two pulleys: when one pulley is active a reverse mode of DC motor can provide force-moments in both directions See also Yoshihiro Imazeki, Kazunori Itoh, Yoshimichi Yonezawa Virtual shape-making system for tactile recognition of the visually challenged. ICCHP'98 Slide 5-18 Bad haptic solutions Illustrations of the vibratory motors (actuators) attached to fingers, data gloves and exoskeleton Fukumoto, M. and Sugimura, T.: Active Click: Tactile Feedback for Touch Panels. In: Proc. of CHI 2001, Interactive Posters, ACM (2001) 121-122 Slide 6-18 goals of the project were: to investigate the perception features of force feedback provided with 1D-cable suspended haptic interface in the task of blind inspection and matching of length of virtual horizontal bar chart columns augmented with wrench and sound feedback Slide 7-18 mock-up of cable-suspended interface comprises: USB Graphics tablet and haptic unit providing cable tension at the particular moment and location the unit consists of DC motor, pulley, and reduction gear (ratio 21:1), to transform with a minimum lost the torque of about 6 mN/m into tractive force of 6 N lead to stylus Slide 8-18 screenshot of the testing software Slide 9-18 experimental setup the tractive power (Psd) was counted as follows: equation: Psd equal to Fsd times L times fraction T1 devided by T1 plus T2 Fsd- tractive force applied to stylus L - length of chart column T1- period of time when DC motor was switched on T2- period of time when DC motor was switched off subject anchors the hand on the desk when grasping a stylus and following the tractive force lead to a stylus to evaluate length of chart columns Slide 10-18 experimental procedure 8 sighted right-handed paid volunteers participated in study no subjects reported any loss or intolerance of haptic perception subjects were blindfolded and wore mask to prevent an access to visual information subjects were asked to leave the hand relaxed while grasping the stylus and to follow the force moments provided with the device without applying strong resistance to the wrench move performance was evaluated in terms of the number of repeated inspections to detect twin chart columns with similar length, and the task completion time required to perform the chart inspection Slide 11-18 parameters of virtual chart columns used Table 1. The parameters of the virtual horizontal bar chart columns used The exploration of the chart was accomplished with chart columns having Lmin value of 5, 10 and 13 mm. Seven different twin-column to non-twin column length ratios were varying on their values in dependency of Lmin and Krnd Slide 12-18 results: repeated inspections per column illustration: Fig. 2 in the paper the average number of repeated inspections made by subjects to explore chart columns in the inspection phase of the match game The average number of repeated inspections increased when twin-column to non-twin column length ratio was more than 4 to 5 and varied from 2.31 to 3.75. The number of repeated inspections varied from 4 to 6 when twin-column to non-twin column length ratio was less than 4 to 5 and Lmin was equal to 10 mm. When column length ratio was more than 4 to 5, the variation decreased to 2.25 - 4. The number of repeated inspections varied from 3.7 to 6.5 when twin-column to non-twin column length ratio was less than 4 to 5 and Lmin was equal to 5 mm, and increased to 4 - 7.5 when column length ratio was more than 4 to 5. Slide 13-18 results: task completion time illustration: Fig. 3 in the paper the mean twin-column detection time in the inspection phase of the match game The mean task completion time varied from 4 to 8 s when twin-column to non-twin column length ratio was less than 4 to 5 and L min was equal to 13 mm. When twin-column to non-twin column length ratio was more than 4 to 5, the mean task completion time slightly decreased and varied from 3 to 7 s. The subjects were required twice more time, of about 12 - 17 s, to explore the chart when twin-column to non-twin column length ratio was less than 4 to 5 and Lmin was equal to 5 mm. Slide 14-18 results: scanpaths recorded illustration of the scanpaths (screenshot) 5 of 8 subjects did not continue exploration of rest part of chart once upon a short chart exploration they were sure which columns were twins Slide 15-18 results: scanpaths recorded illustration of the scanpaths (screenshot) 4 of 8 subjects explored all the chart columns one by one before they became sure which chart columns were twins Slide 16-18 results: scanpaths recorded illustration of the scanpaths (screenshot) all subjects explored all columns one by one for prolonged period of time in case when the twin columns were weakly differed from each other Slide 17-18 Conclusion the players made only 1.5 - 3 repeated inspections and were needed only 4 - 8 s to accomplish the chart inspection the players resulted in lowest error rate of 1.2% when twin-column to non-twin column length ratio was less than 4:5 and minimum length of the column was 13 mm Slide 18-18 Prospects the positive experience acquired within simulated gameplay conditions with the use of implemented cable-suspended interface can be applied in developing novel didactic tools for training blind students in estimating linear dimensions of the simulated objects Thank you for your attention