2 December 2013

Ryan Eubank
Tactical Defense Systems

Unmanned Aerial System Development Challenges: Small Scales and Open Architectures

Abstract
Unmanned aerial systems (UAS) continue to proliferate year after year into both wider development and broader application. They play increasingly significant roles in military and civilian business, and infiltrate the collective commercial, scientific, and social psyche of the modern world. Despite global attention and expansive research expenditure, there remain a number of fundamental challenges to keep pace with their expanding roles in the development and deployment of emergent UAS technologies. Ultimately, new approaches are required to improve interoperability and transfer new technologies into operation. This talk will present a broad vision for a common UAS research and development framework to facilitate and shape the incubation and deployment of robust and innovative technologies. The presentation will provide an overview of several efforts at MIT Lincoln Laboratory, including ongoing collaborations with MIT campus, that are focused on open architectures, extensible system composition, and rapid prototyping concepts to facilitate and accelerate UAS technology propagation. Architectural and compositional concepts will be introduced for both hardware and software in the framework of previous UAS capstone design courses at MIT, with special attention paid to the specific challenges of micro-UAS development.

18 November 2013

David Kong
Bioengineering Systems and Technologies

Metafluidics: An Open Repository for Microfluidics

Abstract
Microfluidic, or “lab-on-a-chip” instrumentation has the potential to be a core enabling technology for research areas ranging from synthetic biology to miniaturized electronics and optics. While thousands of microfluidic devices with myriad applications have been developed, a significant obstacle remains the slow spread of these technologies amongst communities that would greatly benefit from their use. Microfluidics are typically not easy to make or use, and researchers are generally unable to leverage the designs and hardware of other groups. To help address these issues, we are developing Metafluidics, an open repository of microfluidic device and hardware designs for microfluidics.

Through Metafluidics, we hope to build a community of developers (microfluidic engineers) and users (e.g., synthetic biologists) to share and remix designs, thus making microfluidics more accessible to researchers of all types, from students just learning about miniaturization technologies to cutting-edge innovators.

4 November 2013

Dan Schuette
Advanced Imager Technology

Image Sensors to Enable Fast Autonomous Flight at Low Altitudes and in Complex Environments

Abstract
Fast autonomous flight in the complex environment found at low altitudes is a challenging problem for small
UAVs.  Practical considerations severely limit the sensors and processing power that can be supported. Recently, there have been some exciting demonstrations of fast autonomous flight using active sensor systems. While impressive, the active sensor systems used in these demonstrations are large and have power requirements more than an order of magnitude higher than what a passive system should be able to achieve.

To enable operation on smaller platforms and for longer missions, MIT Lincoln Laboratory is developing a passive image sensor that will derive scene 3D structure from the apparent motion (optical flow) created by the aircraft's flight. This sensor integrates an on-chip optical flow processor with a photon-noise-limited image sensor imaging 1,000 fps to enable optical flow calculation for structure-from-motion estimation in real time.

21 October 2013

Gene Itkis
Cyber Systems and Technology

Extracting Cryptographic Keys From Biometrics

Abstract
All cryptographic applications require some secret keys for users. Where can the users keep these keys, or reliably re-generate them? This seminar explores different options, one of which is to derive the keys from biometrics. The theoretical approach is to use fuzzy extractors, a cryptographic construct developed within the last decade. Applying this approach in practice, however, presents a number of challenges. We have investigated some of these on the example of iris biometrics. This talk will outline the main challenges that we encountered in our investigation. This is joint work with V. Chandar, B. Fuller, J. Campbell, R. Cunningham, and J. Borgstrom.

7 October 2013

John Chiaverini
RF and Quantum Systems Technology

Large-Scale Trapped-Ion Quantum Processing for Enhanced Computation, Simulation

Abstract
Quantum information processing promises impressive gains in solving computational problems relevant to elds as diverse as cryptanalysis, precision measurement, tailored-material design, and drug discovery. Signicant technological challenges remain, however, before processing at the large scale required to address these problems is enabled. A leading candidate system for scalable quantum processing is a collection of trapped atomic ions manipulated by electromagnetic elds; the long internal-state coherence within each ion and the strong Coulomb interaction between ions provide long-lived quantum bits (qubits) with controllable multi-qubit interactions. Operational error rates in few-ion systems have been shown to be low compared to other methodologies, but for scalability, these error rates must be reduced while simultaneously increasing operation speed and manipulating large numbers of individual qubits. Toward these goals, I will describe recent studies of the basic limitations of ion processor performance due to trap surface properties, as well as methods to address these limitations. I will also discuss progress toward working with many addressable qubits, in particular techniques for rapid, pure loading of ion trap systems and development of 2D trap arrays for large-scale processing.