EVL is the Electronic Visualization Laboratory in the Department of Computer Science in the College of Engineering at the University of Illinois at Chicago in Chicago, IL. EVL is an interdisciplinary graduate research laboratory that combines art and computer science, specializing in advanced visualization and networking technologies. The laboratory was founded in 1973. It is a joint effort of UIC's College of Engineering and the School of Art & Design, representing the oldest formal university collaboration between engineering and art in the United States, offering graduate degrees in electronic visualization (MFA, MS, PhD.)

EVL has a variety of virtual reality devices which we have developed over the last 15 years making use of stereoscopic projection, head tracking, and tracked controllers. Currently our CAVE(tm) is being refurbished, but we have a single-screen C-Wall, the 4-megapixel ImmersaDesk 4, and the PARIS which is a drafting-table sized display with a haptic SensAble PHANTOM as an input device. We have also developed low-cost non-tracked stereoscopic displays such as the GeoWall (www.geowall.org).

We have several LCD tile display systems including the 11x5 100-megapixel LambdaVision display driven by a 28 node Opteron cluster with 1.5 TB of local storage, the 24-megapixel LambdaTable, and a several single-PC based tile displays.

We are also working on combining these technologies into high-resolution auto-stereoscopic tiled displays (stereo viewing without special glasses) such as the 7x5 Varrier driven by an 18-node dual Xeon cluster, and its desktop version the Personal Varrier.

EVL has a persistent 10 Gigabit Ethernet (GE) connection to the University of Washington in Seattle and the University of California in San Diego via its own private wavelength on the National LambdaRail infrastructure. EVL works with Argonne and Northwestern University to operate the StarLight facility in downtown Chicago, a 1GE and 10GE switch/router facility as a proving ground for network services optimized for high-performance, large-scale national and global applications to Europe and Asia. EVL uses these facilities to develop new protocols that take advantage of these networks and investigate collaborative sharing of high-resolution video and scientific datasets.

Here are some ideas for Milan Thesis work. There are many more possible ones, but this may give you some ideas.

• Intelligent avatar (head, head and upper body, or full body) that you can talk to that will answer your questions (might be a person at a museum that tells you about the exhibit.) This would involve using motion capture data and physics to animate a human head or body, voice recognition (and perhaps camera recognition) to listen to the user, artificial intelligence and some database work to know how to generate an answer, and then speech synthesis for the avatar to talk back. Doing all of this is a big hard problem, but it would be possible to do a part of this.
• Whole city/driving simulator with traffic, lights and pedestrians. We have done smaller driving simulators but not one in a dense area. There are some nice city models available for purchase but they are static. Automobile and pedestrian models are available. It would be nice to create a dynamic cityscape for people to drive through. This would involve creating a large number of agents (people on the street, people in cards), and animating the elements of the city that are near the driver. Maybe this application is collaborative between a 'driver' at a large screen PC and a pedestrian in a VR device, where they both share the same city.
• Evacuation simulation. We are working with the City of Chicago and Argonne National Labs on how to model the evacuation of a tall building in downtown chicago, or even the evacuation of the entire downtown area in the event of something bad happening. Whereas the driving simulation is interactive, this larger set of agents would be run as a simulation on large cluster computers to generate visualizations in 10 or 15 minutes that could be played back.
• Dynamic data overlays for our interactive table. We currently have interactive maps on our 6-tile table. We could use this for maps of Chicago / US / Italy / Europe / World. We can currently zoom in/out and rotate the map. It would be good to add more interesting dynamic data to this map such as: weather, traffic, crime, earthquakes, floods, disease outbreaks. This would involve writing code to access the on-line data, convert it to a usable form, apply it as a user-friendly layer onto the maps on the table, and allow users to query for more information.