Three technologies developed by MIT Lincoln Laboratory were named recipients of 2015 R&D 100 Awards at the 53rd Annual R&D 100 Awards and Technology Conference, held at Caesars Palace in Las Vegas, Nevada. These awards are given annually by R&D Magazine — a publication for research scientists, engineers, and technical staff — to recognize the world's 100 most innovative technology products introduced during the year. Editors of R&D Magazine, aided by a panel of judges with expertise in diverse technical fields, selected the winners from among more than 300 finalists, including 11 laboratory-developed technologies, which were announced in July.
Since 2010, Lincoln Laboratory has received 26 R&D 100 Awards, including two R&D 100 Editor's Choice Awards, which are given to the top three technologies from among the year's 100 selected by R&D Magazine.
The three award-winning Lincoln Laboratory technologies are
- Platform for Architecture-Neutral Dynamic Analysis, an open-source, plug-in software analysis framework that enables computer engineers to observe code as a program executes so they can understand and mitigate vulnerabilities or faults in the code;
- Self-Defense Distributed Engagement Coordinator, an automated decision support tool that guides naval personnel on how to efficiently allocate resources in response to anti-ship missile threats; and
- Video Content Summarization Tool, a software application that creates summary views of long-duration surveillance videos so analysts can quickly identify activity of interest.
Simplifying the analysis of complex software systems
The Platform for Architecture-Neutral Dynamic Analysis, or PANDA, provides a flexible plugin-based framework for analysts to quickly develop instrumentation to help answer questions on how software behaves. The platform was designed to facilitate the reverse engineering technique known as dynamic analysis, in which analysts study software as it executes on a system.
A novel record-and-replay mechanism allows analysts to capture a recording of all software executing on a system and replay that recording an unlimited number of times. To gather information about the software during replays, analysts can implement PANDA's software-analysis plugins, more than 40 of which have been developed by Lincoln Laboratory researchers, university collaborators, and the open-source community. With each replay, the plugins construct a deeper understanding of the important aspects of system execution, from determining what specific processes are executing on the system to tracking specific pieces of data as they flow around the system. Using this approach, analysts can iteratively build knowledge about dynamic software executions. This knowledge can be used to improve the security of software products, to investigate known or unknown software vulnerabilities, and to revive legacy code in modern environments.
Improving the allocation of fleet antimissile resources
The Self-Defense Distributed Engagement Coordinator (SDDEC) manages a fleet's antimissile resources (missiles and electronic warfare technologies, such as interceptors, signal jammers, and decoys) and makes intelligent, real-time decisions regarding their usage as missile threat scenarios unfold. A novel algorithm allows SDDEC to assess critical factors — the location of threats and other nearby ships; the size of the ship's resource inventory; the interactions between ships, threats, and responses; and temporal and spatial variables — and recommend a response strategy that efficiently allocates resources to defeat incoming missile threats.
For each individual ship, the SDDEC develops an initial threat-response recommendation on the basis of the capabilities and costs associated with the ship's resources, current threat trajectories, the ship's value to the mission, and fleet-wide risk-tolerance levels. These initial recommendations are dynamically updated as operators across the fleet share information on threat conditions and apprise each other of their plans. Although designed for fleet defense, SDDEC could be customized for other emergency response situations, efficiently allocating resources such as vehicles, personnel, and supplies.
Speeding up the video review process
The Video Content Summarization Tool (VCST) is based on a novel activity-mapping technique that temporally maps all pixel-level motion detected in long surveillance videos onto a condensed, composite visual clip. Unlike other video-synopsis techniques, VCST does not rely on object tracking, which can be an error-prone process for scenes containing partial occlusions or intersecting motion patterns. Instead, VCST uses a mapping scheme that produces a summary clip containing many instances of foreground motion overlaid onto a static background. This mapping scheme guarantees that motion continuity and activity co-occurrence between the original and output videos are preserved, leading to a pleasing viewing experience for analysts.
Two slider bars on the graphical user interface give analysts instantaneous control over the time-compression ratio (i.e., length of synopsis video) and motion-sensitivity threshold. The ratio can be adjusted to move from sparse to dense activity representations of scenes; the threshold can be set to capture meaningful activity while avoiding oversensitivity to motion clutter, such as that from illumination changes. By clicking on a specific piece of activity in the summary clip, analysts can jump to the original video frames containing that activity to examine it more closely. The software was designed with these customizable settings to support interactive video browsing.
