Ianarcher’s Weblog

Ok an update on the development of the system for eye-to-hand coordination with the mobile haptics equipment. Well it is becoming apparent that the possibility of developing a set up like this (with present mobile kit) requires a lot of configuration to be made, which is unique to the user at the time. What I mean by this is that the present set up (both physical and software based) means that depending on the user they can sit down or stand-up, or even stand off center to the workstation. This would be OK in a standard set-up but as this system attempts to line up the eye, hand and workstation (Haptics model) than these all are crucial points to be determined.

Apart from this there are other problems, to separate the rending of the Haptics from the Visual when using the HL API means that the scene has to rendered for the screen from one view point and then the haptics have to be rendered from another. As well as doubling the processors work load it bring the difficultly of lining the two up and getting variables like perspective and scale correct.

One way I suggest to improve this if not solve these problems is create a initial configuration system so as a new user starts using the system they would be asked to perform task(s) to determine there standing position and present view point. This is the kind of idea that Mobile touch-screen systems and even game systems use when using devices and systems that have these type of configuration needs.

I am not going to develop the present test system any further at present, this decision was taken not only partly because of time constraints and demands from other projects but because I feel that there is no real demand for this at the moment. The brain compensates for the division in the present system so well that I didn’t realise this was even an issue until I took the time to study how my hand was moving the Haptics device without looking at the screen.

I have started working on a way to coordinate the users hand and eye with the mobile haptics equipment, to understand this you first need to understand the mobile haptics equipment.

The mobile haptics equipment in question is packaged by SenseGraphics, it is made up of a shape 3D laptop a SensAble Phantom Omni and a custom built stand that makes it all work (full specification).

As you would expect the laptop has OpenHaptics API installed but as well as this SenseGraphics have installed their personal API called H3D. H3D used the OpenHaptics HL API combined with OpenGL to render graphics both visually and haptically, but the clever part of the API is the file format called H3D which is based on the XML web graphics format X3D.The H3D API comes with a handy H3D file loader; this is where the issue arises. It is very good at loading and display haptic models in fact it the best I’ve seen, but as with everything it is not perfect. The mobile kit setup allows the user’s hand to be semi visible through the glass where the model is projected, so if the model and your hand matched up it would be even more impressive.

David has asked me to see if I could look into the possibility of updating the system to do this and if not what are the major pit falls. So to the system I’m working on, it also uses the OpenHaptics HL API with OpenGL to do graphics but doesn’t us the H3D format as it is only a test system and we do not know yet if it is even possible, if it is possible then i would take the time to integrating it with the H3D format and the loader.

The Virtual Gallery of Contemporary Fine Metal Work (VGCFM) is a multimedia DVD ROM that User-Lab is creating for a research output to the REA, they previously created a CD ROM version for Jewellery. This time there will be Video included in the DVD ROM, this is where I come in. They have already contracted out the creation of a Database for input of Artist Details and One for Object Details to two developers but now I have arrived it falls to me to create one for Video. As well as this I am also responsible for the up keep of the other two and they are developed in different languages and using different approaches.

The first one is for Artist and is developed in ASP.net with a MS SQL Server at the backend, the second is written in PHP also attached to the same Database but this time for some reason is not relational so the relationship of keys between tables is purely keep and managed by the PHP code.

If that wasn’t enough languages I have decided to create my Web App for the Video database in JEE using JSP and JSF to drive the front end. The reason for this is partly because I want to stretch my legs in JEE but mostly it because JEE is quickly becoming the standard for web applications and the Server that I am deploying to already have JEE Application server installed (Tomcat). I will also be just adding new tables to the existing Database for but unlike the Object PHP system mine will definitely be Relational and Normalised to at least 3^rd normal form.

Look-And-Feel, the look of the video database front end will be minimalist to say the least as it has been made clear to me that this system unlike the other two will not be used by anyone outside of the team working on DVD ROM and the time I would spend on making it look good could be used a lot more wisely on the other jobs mounting up for me to do.

As part of my duties as the Software Engineer for User-Lab I am responsible for helping Dr. David Prytherch with his haptics research. I am responsible for the technical aspects of his projects and for helping decide on what directions to take depending on what is possible with the haptics equipment.

I have never come across haptics in a practical sense before, I have used game controllers and systems that utilise haptic technology but have never developed for those systems. This fact if I’m honest fills me with a little dread as I can see the complexities of this already! And I’m only contracted a day week to work on this as User-Lab will require me to work on completely different types of applications and systems for the most part.

David has mentioned areas of interest and possible applications that would be useful to develop, these include:

• A way to Calibrate the Mobile Haptics kit to a line the user’s hand to the interaction
• The ability to take any 3D model and make it usable with the Haptic equipment
• Combining the Haptics equipment with the 3D screen we have as well

These task are achievable I’m sure just not sure how quite yet or which one to tackle first. Think there will be many a late night working one this…