IEEE Haptics 08

-- MarkCutkosky - 06 Oct 2007

• paper116.pdf: The Touch Thimble: Providing Fingertip Contact Feedback During Point-Force Haptic Interaction, Katherine J. Kuchenbecker et al.

• • While I think the paper itself is well written and clear, I'm not 100% sure I understand the purpose, I was left with an "ok...so what?" type of response after finishing the paper.
    • Early on, when discussing how they chose some of the design parameters, they spend a good half page+ describing a quantitative method of determining stiffness and spacing of the thimble...and then at the very end they say they didn't end up actually using that method at all. If they do not use the method, why take the time to describe the method? I would rather them mention that they are aware of the method and reference it, but then go into more detail of how they qualitatively decided those parameters.
    • It's not clear to me whether the subjects performed well with the thimbles because the devices worked well or if the task was too easy. I'm more inclined to believe the task was too easy. If subjects were close to perfect in surface recogniztion for both cases, then what's the point in comparing them? (If I were to use the original thimble and subjects had perfect surface recognition percentages, then I would not have strong motivations to design another thimble...why perform build another thimble to compare to the original thimble when it worked well to begin with?) While I understand there are aspects of the original thimble that make it less than desireable for certain tasks, which likely motivated the design of a different thimble, I don't follow the thought process of deciding upon this particular user experiment to test the validity of the new design.
    • Some minor comments
      • This is a minor comment that I'm sure didn't REALLY influence the results, but the authors state that to reduce any bias towards the two different designs, they referred to them as the "white thimble" and "black thimble" and I'm not so sure that was the most appropriate labeling to reduce bias...perhaps they can re-label them in the future...
      • In the results section, they state that "as depicted in Fig. 5, most individuals could usually identify the surface within 10 seconds" but when I look at Fig 5, it seems that most subjects could identify the surface within 2 seconds...it just wasn't obvious to me that the "10 seconds" was likely referring to outliers. To me, that particular description of the figure and actual figure didn't really match well.
    • Basically, I think it's a well written paper and the design of the thimble is presented clearly, however I don't know that the results of the paper are that interesting. Especially given that the only parameter which seemed to differ was recognition time, and the new design had slightly worse performance. Actually, they state that the device was developed to allow users to feel breaking/making fingertip contact, so why did they test surface recognition? Also, they state that the thimble didn't fit each subject the same way- wouldn't that effect the subject's performance/qualitative preference for one over the other? I just think there needs to be just a little more. This would have all been good pilot experiment data/results that could be used to develop better design/user studies, but there needs to be something more conclusive.

paper132.pdf: 1DOF Sensor and Display system of Haptic and Temparature Sensation Hiroaki Yano et al.

• paper132_YLP.txt: Comments from Yong-Lae
• paper132_SCR.txt: Comments from Seok Chang

paper206.pdf: A Multi-criteria Design Optimization Framework for Haptic Interfaces, Ramazan Unal et al.

• Unal.txt: R. Unal
• paper206_YLP.txt: Comments from Yong-Lae

paper232.pdf: Toward Developing a Velocity Controlled Tactile Impact Display, Brian T. Gleeson, William R. Provancher

• Karlin: I don't have many comments about this paper. It was clear and well written. The only question I had was regarding the force range that could be displayed through the contact block, as well as what range of velocities that are possible. But other than that, I thought it showed promise and the design/electronics were described nicely.
• Jason: Overall, it seemed to be pretty good for a 2-pager, mostly describing the design of the impact thimble. The idea seems solid to me as well as the implementation. Few data were presented as expected in this type of brief. I had only a few things that I thought could have been spelled out a bit more clearly, depending on space.
  • 1. I was curious how the sensor was calibrated.
  • 2. I could not tell how the impact block was driven.
  • 3. It would be interesting to know the range of impact velocities/forces that the device is capable of rendering. This would give some idea of the object interactions that would be possible with the device.