Click here to go back to the Chameleon Description.

Problems with Chameleon

Seven people (non disabled) from ASEL's Robotics Lab informally tested the Chameleon robot. The purpose of the testing was for them to identify obvious shortcomings of the system to be fixed before consumer testing begins. They were asked to place blocks from one compartment to another, stack blocks, and do a short version of the MSMT. Additionally, some of them tried to draw and/or write with the Chameleon.

Definitions of table headers:

Problem: An issue that a tester mentioned as being troublesome.
Priority: A level assigned to a problem indicating the importance of a solution. 1 - important, 5 - not very important
freq: Indicates how often the problem was mentioned. 1 - not often (1 person), 2 - sometimes (2-3 people), 3 - often (greater than 3 people)
Possible Reason/s: Proposed reason/s explaining why the problem/s exists.
Possible Solution/s: Proposed solution/s to the problem/s.

Tester's Comments on Chameleon

Problem Priority / freq. Possible Reason/s Possible Solution/s
1. In certain configurations, the radial master movement is difficult. 3/1 Kinematic mismatch between the master unit and person's head. Custom make master unit for user.
2. When system is mechanically recalibrated for a particular task, a kinematic mismatch results. 3/1 The user is trying to do tasks wrt the slave, but the master and slave no longer align. -
3. Can not disengage from master in any position without the system going unstable. 1/1 The system is not balanced throughout its full ROM. Hence, the slave drives the master to a position where the feedback loop of the radial motion takes over. *Balance the system throughout its full ROM.
4. Some tasks are difficult without gripper pitch capability or special gripper adaptations. 3/2 Some tasks simply require this DOF. Add this capability (controlled thru what input?).
5. Can not easily reach all items that are easily visible. 3/1 Links are not long enough or slave is positioned too far back. The links can be made longer if system weight can be reduced to keep inertia to a minimum.
6. The master and gripper units are obtrusive and sometimes interfere with `line of sight'. 2/3 - -
7. The master unit is ugly. 4/2 - -
8. The system feels heavy. 1/1 Poor balancing, Too much friction is present in Bowden cables. *Use a cable system with less friction. *Balance the mechanism throughout more of its ROM.
9. Pain at the back of the jaw and in the neck occurs. 1/2 The mouthpiece may be: too thick, too hard, not the correct shape. The load may be too large. *Possibly redesign the mouthpiece.
10. Radial slave motion occurs with just pitch or yaw intended inputs. 3/3 The radial motion is easier than the others, hence any component in this direction may activate its movement. *Change the type of input (position) that controls radial motion.
11. Can not tell when gripper has fully grasped object. 4/2 No force Feedback -
12. Too large of a position dead-band in radial motion. 1/3 The gain acting on the position error is not large enough to draw enough current to drive the motors for small position changes. * Increase gain acting on position error.
13. Time delay in radial motion makes system difficult to control. 2/3 User moves master input faster than slave's radial motion capability. -
14. Difficult to isolate gripper movements. 1/2 User does not have sufficient tongue control under certain conditions. *Don't use all 4 JS motions for control.
15. Drooling 1/2 User can not close mouth because mouthpiece is too thick. Decrease mouthpiece thickness at certain locations to allow mouth closure.
16. Gripper wires get in the way. 3/2 They presently dangle. Reroute
17. Difficult to do fine motions in the radial direction. 2/3 Gain is too large acting on position mapping between master and slave radial motion, gain is too small acting on position error,time delay * Increase gain acting on position error.
18. Difficult to synchronize yaw/pitch motion with radial motion to get smooth motion. 2/2 Control with force feedback vs. w/o, differences in effort required to activate joints. -
19. Forces felt are really damped out. 4/2 System stiffness is not enough. -
20. When head is pitched at a significant angle, yaw motion was very difficult. In general, out of horizontal plane motions are more difficult. 3/1 Kinematic alignment problem. Redesign master unit so that it pitches with the user's head.
21. Difficulty holding mouthpiece in mouth. 1/1 Mouthpiece is too thick; it is not molded to fit the user's dentition. *Redesign mouthpiece
22. Teeth contact occurred only at molars; not in front. 3/1 Same as above. Same as above.
23. Some people needed frequent breaks. 3/2 Too much effort required to power the system, need to break for salivating -
24. The system is tiring to use. 2/2 - -
25. The user needs to move head back too far to comfortably use the system. 2/1 The master's mount is located too far forward. * Translate the master unit forward.
26. As master moved radially in, the mouthpiece`s plane changed wrt the user's bite plane making it uncomfortable to use 1/1 The effective radius' of the user and master are significantly different, poor initial set-up Custom make master unit. Take more care aligning master to user.
27. The user's eyes bothered him when not aligned with the object being manipulated. 2/2 - Calibration needs adjusting.
28. It is difficult to donn/doff the system. 3/1 - -
29. Easily forget which tongue motion corresponds to which gripper motion. 2/3 Non-intuitive mapping, requires learning *Remap control of gripper to something more intuitive.

* Denotes the proposed solution has been implemented or addressed.




Click here to go back to the Chameleon Description.