RIT’s remote-sensing superheroes
Feb. 16, 2012
by Susan Gawlowicz
Follow Susan Gawlowicz on Twitter
Follow RITNEWS on Twitter
Thirty years ago, at the beginning of Don McKeown’s career at Kodak, his mother put a stuffed Smurf in his lunch bag as a joke.
“I’m known in my family as having an extreme hatred of Smurfs. It goes way back,” McKeown says, glancing at the Papa Smurf peering out of a coffee cup on a shelf in his office. “I got to work and opened my lunch bag and there was that stinking Smurf,” McKeown says. “And so he’s followed me for my whole career and wherever I was, I always put him in a spot. He’s been on shelves, hanging on light fixtures; he’s been everywhere.”
What started as a good-natured prank over the years doubled as a subtle reminder to McKeown not to take himself too seriously and to keep his feet firmly on the ground.
McKeown brought his down-to-earth approach with him to RIT in 2001, when he joined the Chester F. Carlson Center for Imaging Science as a distinguished researcher with an expertise in remote sensing and operation management. During the last few years, he has actively promoted RIT’s Information Products Laboratory for Emergency Response as a resource for the disaster management community and the remote-sensing industry. McKeown helped launch the National Science Foundation-funded laboratory in 2009 with imaging science professors Jan van Aardt and Anthony Vodacek, and Chris Renschler at University at Buffalo.
“I think we’re grounded and realistic in what we think we can do,” McKeown says. “We go beyond traditional research. It’s a hybrid of pure and applied. These users want information, not data.”
The Information Products Laboratory for Emergency Response pulls together groups of people who normally don’t mingle. It gives industry members a chance to ask emergency managers what they want and need in a crisis and how they can improve the information products they provide—the detailed digital maps of disaster zones. RIT promotes itself to both groups as a resource for conducting targeted research and development of electronic maps and customized sensor and processing systems. In this way, the laboratory creates a pathway for transferring technology and knowledge from academia to industry and the end-users in the field.
At the heart of the laboratory is a multi-spectral camera system fondly known in the Center for Imaging Science as “WASP.” It began life a decade ago as a prototype designed and built for the U.S. Forest Service as part of the federally funded Wildfire Airborne Sensor Program. Technical tweaks and major overhauls over the years have improved WASP’s ability to capture high-resolution visual and infrared imagery with its four individual cameras.
Jason Faulring knows WASP better than anyone in the Digital Imaging and Remote Sensing Group. The young research engineer’s career at RIT has evolved and grown alongside the camera system that launched RIT into wildfire remote-sensing research and led to the Information Products Laboratory for Emergency Response.
Faulring helped design and build the 300-pound camera system that peers through the hole in the floor of a small plane. His expertise in operating it has taken him to unlikely places—Haiti, after the earthquake of 2010; prescribed fires in Kentucky, Georgia and Florida; and, most recently, to the floods that slammed the southern tier of New York following Hurricane Irene and Tropical Storm Lee.
McKeown, Faulring and their colleagues in the Center for Imaging Science are working toward real time air-to-ground image transmission to deliver information to emergency managers as quickly as possible.
“In the world of applied remote sensing, so many things have to come together—the flight, the sensor, your health, the weather—all these things are stacked against you,” Faulring says. “Fortunately, we’ve been able to work through them by being flexible and able to adapt to ever changing and challenging situations.”
In late August, Faulring sat in a twin-engine Piper Navajo owned by RIT partner Kucera International and focused the million-dollar camera system over the receding Schoharie Creek. After landing in Rochester that evening, he pushed the imagery through a new processing and geo-referencing workflow that finished the task in a mere three hours; the data was then delivered over the Internet to New York state emergency managers in Albany.
McKeown led the team that looked at the more than 1,400 scenes of water overflowing the Gilboa Dam, contaminated plumes and structural damage, extracting areas of interest to highlight for the emergency managers. Kucera and RIT flew the mission at no cost to the state to aid in the effort and to demonstrate the technology.
A week later, Tropical Storm Lee pounded the East Coast, and Faulring was back in the Navajo. This time the New York State Office of Emergency Management and the Federal Emergency Management Agency funded RIT and Kucera to fly to Binghamton to map the cresting Susquehanna River. The slow-moving flood had forced 100,000 people to evacuate from New York to Pennsylvania when the river rose to record levels.
Low clouds caused the crew to switch to the contingency plan McKeown had made for low-altitude operations following the Schoharie experience. The crew covered 23 miles and collected another 1,400 scenes in two hours, including imagery of contamination slicks the state Department of Environmental Conservation found useful. RIT posted the processed images 12 hours after authorization to fly, only six hours after the first frame was collected.
McKeown organized and led a three-hour post-mortem on the Schoharie and Binghamton flood missions in December at the Institute for the Application of Geospatial Technology in Auburn, N.Y. Attendees and online listeners included members of the state and federal Office of Emergency Management, FEMA, the Department of Homeland Security, the Department of Environmental Conservation and the U.S. Geological Survey, which coordinates natural-disaster response.
Dennis Pokrzywka from the New York State DEC Spill Unit pointed to the potential for detecting and mapping unknown contamination slicks. He showed examples of RIT imagery and ground-based images around the BAE Systems plant in Johnson City of a plume of heating oil.
“Finding out what doesn’t work is just as important as what does work,” McKeown notes.
In time, McKeown is confident emergency managers will ask their industry partners instead of RIT to provide remotely collected imagery during disasters, fulfilling one of the laboratory’s objectives.
“We are not in the business of responding to disasters,” McKeown says. “What we want to do is to enable someone in the commercial sector to be able to do that. We want to take what we learn through demonstrations—through occasional participation in disasters—to take those insights and to transfer them over to the commercial sector, which can then be on call 24/7. Now, if there is another major event in New York, is the phone going to ring? It might.”