News & Ideas is a guide to research in science, technology, management, architecture, and humanities and social sciences at Rensselaer. For details or photos, contact Marketing and Media Relations, Rensselaer Polytechnic Institute, Troy, NY 12180, (518) 276-6532, or e-mail us at nasons@rpi.edu.

SURGERY:
Sharing control with a robot

BIOMEDICAL ENGINEERING:
Curves in arteries create risks

ENGINEERING AND ECONOMICS:
No more World Wide Wait

INTELLIGENT AGENTS:
A society of drill presses

LAB-GROWN ARTERIES:
No rejection after surgery

APPLIED BIOPHYSICS:
Animal-free product testing

SURGERY:
Sharing control with a robot

Researchers at Rensselaer Polytechnic Institute are developing an Endobot, a robotic tool that allows surgeons to manually control as much of an operation as they choose and turn the rest over to the robot. The goal is a device that can increase the speed and precision of minimally invasive surgery while reducing the surgeon's level of fatigue.
Now being modified as a result of earlier tests, the prototype features two arms that manipulate interchangeable tools for grasping, cutting, retraction, and stitching.
Unlike robotic devices that rely on techniques such as teleoperation, the Endobot keeps the surgeon in touch with the instruments, controlling them manually and with joysticks and pedals. The doctor can position the Endobot's arms precisely, for example, and then allow them to take over the stitching. One arm holds the flesh in place while the other stitches, finishing the job in seconds rather than minutes.
Sunil Singh and John Wen, researchers with the Center for Advanced Technologies (CAT) and the Department of Electrical, Computer, and Systems Engineering, at Rensselaer are developing the Endobot in collaboration with Dr. Bart Sachs, director of the Minimally Invasive Surgery Center at Albany Medical Center. Last June, Singh, Wen, and Sachs formed a company, Endobotics Inc., that brought the idea to the CAT for development. The company is now patenting the device.
The CAT researchers also are working with Dr. Dana Mears, director of orthopedic surgery at Albany Medical Center, to add drills, reamers, and other tools that would be needed for minimally invasive hip and knee replacements.
Sachs described the technology in March at the annual meeting of the Society of American Gastrointestinal Endoscopic Surgeons in San Antonio.
In minimally invasive surgery, doctors do not cut patients wide open. Instead they make small incisions and send in cameras and tools that can be operated from a distance. This greatly reduces pain and recovery time, but it complicates surgery for the physicians, who can see what they are doing only on a video monitor and who must precisely cut and stitch using instruments located on the ends of long handles.

Contact: Sunil Singh (518) 276-2243, sunil@cat.rpi.edu.

BIOMEDICAL ENGINEERING:
Curves in arteries create risks

Atherosclerosis, an arterial disease that can lead to heart attacks, strokes and aneurysms, develops in particular regions - areas where branches or sharp curves create disturbed blood flows.
Natacha DePaola, assistant professor of biomedical engineering at Rensselaer, is trying to understand how the blood flow causes the changes, basic knowledge that could one day be used to devise preventive measures or cures. She and her collaborators have now shown in quantitative laboratory studies how the disturbed blood flow disrupts communication among the endothelial cells that line the arteries, research that was reported in March in the Proceedings of the National Academy of Science.
Connexins are proteins that make up the communication channels between cells. In the endothelial lining, healthy cells express mainly connexin40 (Cx40), while cells in diseased areas express another connexin type, Cx43.
DePaola's group designed laboratory equipment to study smooth and disturbed fluid flows over a layer of endothelium cells and then observed and measured the disruption in cell communication over time. In disturbed areas, the cells - not getting expected information about the presence of adjoining cells - began to proliferate abnormally.
In the study, DePaola and collaborators at the Institute for Medicine and Engineering at the University of Pennsylvania, identified and measured not only protein expression and function of Cx43, but also changes in its gene expression.
In other studies, DePaola's group has looked at the movement of endothelial cells, which migrate from disturbed areas, and at the basic breakdown in the ability of the endothelium to serve as a barrier. DePaola's work is funded by the National Science Foundation and the Whitaker Foundation.

Contact: Natacha DePaola (518) 276-2170, depaola@rpi.edu

ENGINEERING AND ECONOMICS:
No more World Wide Wait

A practical solution to Internet traffic jams is the focus of research at Rensselaer Polytechnic Institute. The project, which is funded by a $455,000 grant from the National Science Foundation, blends engineering with economics.
Many engineering proposals for reducing Internet congestion fail to consider economic issues that concern users and Internet Service Providers (ISPs), while other remedies propose economic controls that are not workable from an engineering standpoint, say Shivkumar Kalyanaraman, assistant professor of electrical and computer engineering, and Thiagarajan Ravichandran, assistant professor of management in Rensselaer's Lally School of Management and Technology.
Until the congestion problem is solved quality and speed of service will suffer, increasing what has been called "the World Wide Wait," says Kalyanaraman.
The Rensselaer project will be the first to bring together experts in both management information systems and engineering to address the problem of Internet congestion.
"Our ideas were partly inspired by the use of congestion-sensitive prices on a freeway in Irvine, Calif., which succeeds in avoiding traffic jams," says Kalyanaraman.
Charging for service may be necessary, with rates that vary with bandwidth needs and network congestion, says Ravichandran. ISPs must be able to make a profit in order to encourage the infrastructure investments, and the right of access to electronic information by all segments of the population should also be considered in pricing Internet usage, Ravichandran says.
The researchers will develop computer simulations based on a variety of pricing schemes, potential changes in technology, and models of Internet use and human behavior.

Contacts: Shivkumar Kalyanaraman (518) 276-8979, kalyas@rpi.edu; Thiagarajan Ravichandran (518) 276-2035, ravit@rpi.edu.

Imagine a factory as a society in which parts, machines, and material handling systems negotiate with each other - each trying to meet its own objectives and at the same time contribute to the goals of its department and of the company.
Sunderesh Heragu, associate professor of decision sciences and engineering systems (DSES) at Rensselaer Polytechnic Institute, and Robert Graves, DSES professor and director of the Electronics Agile Manufacturing Research Institute at Rensselaer, have been awarded a three-year $426,000 grant by the National Science Foundation to build a computer framework that creates such a system.
Heragu, principal investigator on the project, explains that while a growing number of researchers are exploring ways to use intelligent agents in distributed computing systems, intelligent agents are not being used successfully to control manufacturing environments. Instead, current manufacturing control systems tend to be hierarchical, sending orders from the top or heterarchical, focusing almost entirely on interactions between individual units.
Heragu and Graves plan to build and test a prototype of an integrated system that does both. Their prototype will control the materials handling and machine systems in a factory. Eventually, the concept could be expanded to include sales, customer service, marketing, and other departments in a fully integrated control system, they say.
Their proposed "holonic framework" is based on the biological model of holons, subsystems that preserve and assert their individuality while at the same time functioning as an integrated part of a larger unit.
In the proposed model, one intelligent agent might represent a part that has the goal of passing through the system as quickly as possible while another represents a drill press that prefers to take on high-priority jobs first. Agents for various drill presses would bid on a job, taking into account their own priorities or workloads. The agent for the drill press cell (the foreman) might assign a priority job if no individual drill press bids on it. The system agent (the factory manager) could transfer jobs from one unit to another or assign an overriding priority to a particular part.
The cell and system agents would take into account the needs of the individual parts, machines, and material handling devices, but they would also be programmed to serve the overriding goals of the factory.

Contacts: Sunderesh Heragu (518) 276-6856, herags@rpi.edu, Robert Graves (518) 276-6955, graves@eamri.rpi.edu

LAB-GROWN ARTERIES:
No rejection after surgery

Someday, the arteries used in coronary bypass surgery may be grown in the laboratory, made from the patient's own cells, which the body will not reject. What's more, doctors will not have to remove a vein from the patient in an additional, invasive operation.
To speed that medical breakthrough, computer scientists from Rensselaer Polytechnic Institute have been working with researchers from California, Colorado, and Minnesota to develop a strong, tissue-equivalent artery that the body will accept as its very own.
The plan is to erect a superstructure of collagen on a TeflonTM tube - almost like the steel-beam skeleton of a skyscraper. Then the scientists would incubate some smooth muscle cells from the patient on the collagen superstructure, creating a user-friendly, rejection-resistant artery.
The collagen would eventually dissolve, but the muscle cells of the new artery would thrive and continually renew themselves according to the established structural pattern.
"Because the artery has to withstand intense, pulsing pressures, we have to replicate the ribbed construction and circumferential strength of the natural artery," says Joseph Flaherty, Amos Eaton Professor of Computer Science. Flaherty is developing computer models that can be used to optimize the design and simulate fabrication.
"We may need to grow arteries that are more thickly walled at the ends, where they will be attached because the surgical connections have to be strong," says Flaherty.

Contact: Joseph Flaherty (518) 276-6348, flaherje@cs.rpi.edu

APPLIED BIOPHYSICS:
Animal-free product testing

A Rensselaer Incubator company has commercialized a technology that may take the animal out of animal testing. Ivar Giaever, a Nobel Prize-winning biophysicist, and Charlie Keese, a senior research scientist, have developed the Electric Cell-Substrate Impedance Sensing-ECIS 100TM-which uses electricity to study complex cell behavior.
Giaever and Keese founded Applied Biophysics Inc. (www.biophysics.com) at the Incubator in 1993. Since then, they've licensed the ECIS technology for commercial use and have sold more than 20 systems worldwide. Giaever presented his research at the American Association for the Advancement of Science annual meeting in Anaheim, Calif., in January.
The new electric biosensor offers unprecedented sensitivity and detailed results, says Giaever. Data can be taken as often as every quarter-second to follow a cell's behavior movements. The software runs on a Windows 98 platform and manages all data acquisition, storage, and analysis.
"This method offers a non-invasive technique for testing animal cells," Giaever explains. "By electronically eavesdropping on cells, we can examine and measure the activity of live cells over time. This is an entirely new way of doing tissue culture."
In the cosmetics industry, for example, live animals are used to test products. The ECIS 100 eliminates the need for animals and yields more comprehensive data about toxic effects on cells.
The units are in use in Japan, Germany and Taiwan, as well as in several U.S. universities and biotechnology companies such as Genentech Inc. in South San Francisco. There are four systems in use on the Rensselaer campus.
Giaever and Keese began developing the ECIS 100 system in 1991 with Small Business Innovation Research funding from the National Institutes of Health.

Contacts: Charlie Keese (518) 276-6438, keesec@rpi.edu; Ivar Giaever (518) 276-6429, giaevi@rpi.edu; Applied BioPhysics Inc. (518) 276-2165, info@biophysics.com