Young 1)"A Constraint Equation Algebra as a Basis for Haptic Rendering" This article introduces the mathematical framework for handling constraints. In general , the constraints are combined constraints from several sources. The author define 5 base case for the case of 3DOF rendering system with a single point proxy. By combining these 5 cases , we can determine the final constraint easily. But in general case where constraints can be applied to different points on the proxy , it appears more cases will be needed , this work is in progress. It deals with only simple case in which constraints are applied to only one point. But It seems useful to extend to general cases. ------------------------------------------------------------------ 2) Fuzzy Rule-Based Evaluation for a Haptic and Stereo Simulator for Bone Marrow Harvest for Transplant (Matt) This paper deals with developing a low-cost haptic and stereo simulator for bone marrow harvest for transplant. The feedback sensation associated to each layer of the tissue was modeled from tactile sensation descriptions done by the specialists. Fuzzy rules were used to determine the success of the harvest. No, not immediately relevant. ------------------------------------------------------------------ 3) Haptic Collaboration over the Internet (Andrew) Summary: they "present a distributed architechture for haptic collaboration via the internet". depending on the user's network latency, he/she would be assigned to a different speed group. They discuss their methods of communication between users within a group, and how objects would interact within the group. Possible applications: - to provide ideas on network communication and how to set up a protocol for haptic collaboration via networks. ------------------------------------------------------------------ James 4) Haptic Control of a Simplified Human Model - Goal is to create a realistic human model that can interact realistically with objects in a virtual world. - There is no direct way to produce reaction forces from a kinematic (IK) solution alone. (ie. Collision response, inertia, momentum, compliance, and gravity). - Dynamics is better because it utilizes Newtonian mechanics; forces are already part of the calculation and dynamic equations of motion allow simulation to take place at an interactive rates. - A high sampling/calculation rate must be achieved (1000Hz); otherwise, a washboard-like effect occurs. - Multi-body dynamics realistically simulates the many parts of the body and their forces on each other. - Balance the complexity of the multibody dynamics computations with the accuracy of the voxel (Voxmap point shell) based collision detection method. Interesting, but not very relevant to any of our projects. ------------------------------------------------------------------ 5) Haptically perceived orientation of a planar surface is altered by tangential forces (Sara) Summary: they added force fields tangential to a rotated surface plane & had the force field either oriented towards subject or away from subject to see how it affected the user's perception of surface orientation with no visual cues. -It DID affect the perception, but exact effect varied considerably across trials (not consistant) -(visual cues tend to be more important than haptic cues in general) -Interesting technique of trial w/ both naive and non-naive subject No, not immediately relevant. ------------------------------------------------------------------ 6) Influences of Network Issues on Haptic Collaboration in Shared Virtual Environments: (Gary) When networking haptic systems, issues such as delay and jitter can create an unpleasant user experience. Two such problems are: network delay in communicating object collisions causes excessive rebound at the haptic arm; and variation in the delay that different users experience results in inconsistencies in the shared virtual environment. - large delays cause the haptic arm to enter an object deeper before receiving collision information. this is followed by excessive vibration once the arm rebounds - in a single-user, networked application, delay below 60ms was found to be tolerable Brings up good issues but not relevant for the foreseeable future (since we only transmit position info over the network). ------------------------------------------------------------------ 5) LEM - An approach for physically based soft tissue simulation suitable for haptic interaction. (Matt) This paper presents LEM, Long Elements Method, a new approach for physically based simulation of deformable objects. Advantages: 1) The number of elements used to fill an object is one order of magnitude less than in a discretisation based on tetrahedric or cubic elements. 2) The graphics and the haptic feedback can be directly derived from the elements, and no intermediate geometric representation is needed. 3) The use of static instead of PDE equations avoids all the problems concerning numerical integration, ensuring stability for the simulation. 4) No pre-calculations or condensation are used, in order to enable real time topology changes. Yes, we might need this when we deal with deformable objects. ------------------------------------------------------------------ 8) A parameter space for perceptually stable haptic texture rendering (Andrew) Summary: The author's goal was to quantify the range of the stiffness parameter K where haptically rendered textures feel stable. The surface they used was a one-dimensional sinusoid superimposed on a flat surface. They used two exploration methods, Free-exploration and Stroking. What they found: K is in the range of 0.26 to 0.19 N/mm for A=2.0mm, and L=2.0mm K is generally has a larger range of stability for stroking, than free exploration. Possible Applications for VR Project: - to determine stable surface stiffnesses constant K. ------------------------------------------------------------------ James 9) Pressure Masks for Point-like Contact with Elastic Models - Possible to consider the contact area to be truly a point for rigid models, this is not possible for elastic models, as infinite contact pressure can lead to various inconsistencies. - The pressure mask approach is particularly effective for haptics when used with linear elastostatic models with precomputed Green's functions. o These Green's functions can be precomputed for a particular boundary value problem, and then extended as reference to rapidly compute components of these boundary value problems at runtime. - Mask is created from a mesh of polyhedrons (in this case... triangles). o Mask can then be used to alter an elastic model from a single point. (example: a spherical mask can be used when pulling out a point of an elastic model, the mask would cause chunk, that is defined by the mask's radius of influence, of the model to be pulled out). Interesting. Perhaps relevant to the modeling project. ------------------------------------------------------------------ 10) Scaling Issues for Teleoperation: (Gary) When scaling the movement of a haptic arm to a smaller or larger workspace, the feedback forces should also be scaled to prevent excessive stiffness. If done correctly, scaling can be an effective technique to compensate for a limited workspace. - the feedback forces should be scaled by the same factor as the arm movement to retain a natural feel - scaling is more of an issue in teleoperation than in networked collaboration Not relevant to our project. ------------------------------------------------------------------ James 11) Using Haptics and Sound in a Virtual Gallery - Describes an application using a combination of haptics (Phantom), graphics(VRML2), and sound. Description of an application, no theory or implementation discussed. Not very relevant. ------------------------------------------------------------------ 12) The Visual Haptic Workbench (Sara) Utah built a similar setup to ours, w/ addition of a data glove for non-dominant hand, for scientific visualization work. (Used different headtracking, too.) - Co-located visual (stereo) & haptics, didn't mention problems. - Moved emergency stop to be a foot-switch. - Investigated calibration & co-registration (no real details, though) - 5 processors. Do we have enough processing power? - processes communicate via shared memory. Yes, generally relevant to our project.