Wednesday, May 25, 2011

Humanitarian demining: An interview with Lahiru Jayatilaka

Lahiru Jayatilaka is a researcher at the Harvard School of Engineering and Applied Sciences. He leads the project called PETALS, which aims to create visual support devices for use by humanitarian deminers. We recently had the opportunity to interview him.


HCI for Peace: What has inspired you to conduct research on computer-based solutions to clear landmines?

LJ: Thrishantha Nanayakkara introduced me to the severity of the landmine problem during my Sophomore year at Harvard. After a couple of conversations with him, I was moved by the humanitarian urgency of the problem, and simultaneously shocked at the state-of-the-art in the domain. I, like most others, could not wrap my mind around the fact that most landmine clearance occurred with humans using metal detectors and hand-held probing tools. In these times of ‘high-tech’, it seemed unjust to have a fellow human being manually finding and removing buried explosives--a missed audio signal seemed to be the difference between life and death. Furthermore, I found it hard to accept the reality that countries such as Vietnam and Cambodia had remained contaminated with landmines since around the 1970s. Why was landmine clearance so slow? Why did we have to risk human life to get rid of these post-conflict threats? These ponderings, along with a keenness to apply my skills as a Computer Scientist to interesting, challenging and socially relevant problems, were some of the driving factors behind my decision to enter this space.


HCI for Peace: Can you give us a brief overview of the PETALS project?

LJ: The PETALS (Pattern Enhancement Tool for Assisting Landmine Sensing) project is a recent research initiative at Harvard University that aims to develop assistive visual devices to improve the safety and efficiency of metal detector-based humanitarian demining. PETALS is motivated by two fundamental characteristics of humanitarian landmine clearance. First, metal detectors are, and will for a very long time remain, the primary detection tool for finding landmines. Second, metal detectors are of limited effectiveness, as modern landmines contain only minimal amounts of metal, making them difficult to distinguish from the ubiquitous but harmless metallic clutter littering post-combat areas. In response to these challenges related to detection, the PETALS system enables deminers to visualize ‘metallic signatures’. A metallic signature is simply the outline of an electromagnetic field belonging to a buried metallic object. Given that a metallic signature is roughly proportional in size and shape to the buried metallic object, visualizing metallic signatures provides more information to the deminer about the nature of the object buried in the ground, thereby, enabling better detection decisions.


HCI for Peace: What are your plans for the future of PETALS?

LJ: We are currently prototyping the PETALS system with both novice and expert humanitarian deminers at a demining training center. These preliminary field studies have helped us gather ecological support for our experimental findings, and also ascertain feedback about eventual system design. We hope to continue with these field studies during the coming months, while simultaneously transitioning PETALS to a field-ready system. If our plans proceed as expected, we hope to deploy PETALS commercially in the not-so-distant future.