|
|||||||||||||
|
|||||||||||||
Imagine the ILC is up and running. Electrons and positrons collide happily and scientists are taking data. Suddenly there's a problem with one of the laser wires. All experts are at a meeting on a different continent, but the problem needs to be fixed immediately. Difficult? Not when there's a Multipurpose Virtual Lab in place. High-speed, high-resolution cameras would allow the faraway experts to look at the fault, a web-based portal would let them access the controls and tools of the system with a simple "single-sign-on" procedure. And if you think this is far-fetched, think again: the first prototype was tested during the last week of August. And it works! "I was in a hotel room in Berkeley, California, and was able do all the things you can normally do in the HERA control room and nowhere else," explains Ferdinand Willeke, one of the developers of the new tool. His partners in the project, the 20-strong collaboration (an Italian-German EUROTeV workpackage), were in a meeting at DESY in Hamburg, Germany. After 1.5 years of programming work on the Global Accelerator Network Multi-Purpose Virtual Laboratory (GANMVL), the international team set up their prototype for the first time. "It's a great success. Naturally we found a couple of things that need fixing, but that will help us improve the system," says Matthias Kasemann, leading scientist at DESY. The next step is to offer it as a tool to the whole community for testing and use the next guinea pig for the virtual lab is the Calice collaboration. They run a test of their calorimeter at CERN in October and will try to control it remotely. However, the virtual lab is not just about remote operation. In principle it is already possible to run a control room remotely. This system is radically different in that it takes into account the human aspect of teamwork around the world. How do you get a virtual team to be as efficient as a real one? Why do we have problems working together over distances when it seems so easy when we're all together? For these questions, the physicists and computer scientists sought the help of two psychologists who specialise in ‘human-computer interfaces'. "The biggest obstacle is trust," psychologist Markus Hodapp from Mannheim University sums up the first results. "In a normal working environment you know your colleagues and have a estimate of the level of trust you have in them. In the virtual team, you sometimes have to trust your opposite blindly – people are not happy with that." The psychology team from University of Mannheim, Germany and University of Udine, Italy has conducted several interviews and a survey with the team members. Preliminary results show that changes in the working environment, trust in the virtual colleague and possible disadvantages in the job situation seem to be the biggest problem. And lack of discipline, as we have all experienced. They also evaluate the usability of the tools to make them as human-friendly as possible. But how do you build up trust? Apparently there's no avoiding face-to-face meetings. "This is especially important in the start-up phase of a project," says Hodapp. Videoconferences help, because a lot of information is transmitted non-verbally in body language, facial expressions or gestures, and seeing them is already a step up from telephone conferences. "Ironically, in our virtual team working on virtual we have the same problems as everybody else," says Hodapp. Informal things – like virtual coffee breaks, for example – seem to be another solution because social interactions help build the needed trust. The implications of a working virtual control room are enormous. Not only would it make it possible to run the ILC nightshift-free for 24 hours with control rooms in the three regions handing over duty when the next region wakes up. It might revolutionise virtual collaboration in completely different areas, just like the world-wide web developed from a particle physics tool into a thing the world couldn't be without. "Particle physics has often played a pioneering role in developing new tools. At the moment there is no system that takes care of all the communication aspects of global conferencing, but I am sure this will come," predicts Willeke. "We should tackle this." The other aspect in which GANMVL is different is its mobility. While other systems only allow access to electronic controls, this one lets the experts operate and fix real hardware. "You need only minimal support at the machine, like an extended arm," says Willeke. Technical equipment can be handled and repaired, while the expert does not have to get into a car or worse, a plane to fix the bug. Even if there are normally only about five people in a control at any time, the accelerator relies on a team of about 150 experts on the various different pieces of hardware. It is these people who need to keep a close watch on their system. At the ILC, they could be spread all over the world. -- Barbara Warmbein |
|||||||||||||
© International Linear Collider |