Visual Sensor With Resolution Enhancement by Mechanical Vibrations
Oliver Landolt, Ania Mitros, Christof Koch

The resolution of both biological and man-made vision systems is limited by the finite spacing between receptors. This limit can be overcome by applying continuous low-amplitude vibrations to the image or taking advantage of existing vibrations in the environment. Some animals rely on this principle for improved visual resolution. We are applying it to a novel CMOS visual sensor to increase resolution and decrease fixed pattern noise. (full report)

Micromachined Gyroscope Using Operating Principles from the Fly's Halteres
Oliver Landolt, Zhigang Han, Christof Koch, Yu-Chong Tai

We are developing a surface micromachined 2D angular velocity sensor -- also known as gyroscope -- with the intention of minimizing power consumption. By using a detection principle inspired by the fly's haltere system, we expect our sensor to tolerate a higher noise level than previous designs for detecting the direction of the axis of rotation, thereby enabling a significant reduction of supply voltage and power consumption. Another feature is that the mechanical structure will be fabricated with a material called parylene using a novel technology developed in-house. The target application is flight control in extremely small air vehicles. (full report)

Guiding a Robot with an Analog VLSI Motion Sensor Based on the Visual System of the Fly
Reid Harrison, Christof Koch

Sensing visual motion gives a creature valuable information about its interactions with the environment. Flies in particular use visual motion information to navigate through turbulent air, avoid obstacles, and land safely. Mobile robots are ideal candidates for using this sensory modality to enhance their performance, but so far have been limited by the computational expense of processing video. Also, the complex structure of natural visual scenes poses an algorithmic challenge for extracting useful information in a robust manner. We address both issues by creating a small, low-power visual sensor with integrated analog parallel processing to extract motion in real-time. Because our architecture is based on biological motion detectors, we gain the advantages of this highly evolved system: a design that robustly and continuously extracts relevant information from its visual environment. We show that this sensor is suitable for use in the real world, and demonstrate its ability to compensate for an imperfect motor system in the control of an autonomous robot. The sensor attenuates open-loop rotation by a factor of 31 with less than 1 mW power dissipation. (full report)

A 2-D Change Detection and Postitioning System Analog VLSI
Theron Stanford, Christof Koch

We are designing analog CMOS chips which will extract information about moving objects such as their relative size, position, and velocity. We are using analog circuits because of their high-speed real-time performance. Immediate applications of this type of chip include electronic security systems, on- or off-vehicle sensors for intelligent transportation systems and target detection systems. (full report)

A CMOS Imager with Focal-Plane Computation for Feature Detection
Alberto Pesavento and Christof Koch

We designed and tested the first CMOS imager with analog VLSI focal-plane computation for feature detection. The chip implements a modified version of the Tomasi-Kanade algorithm that is suitable for integration in a compact analog VLSI chip. The chip has an array of 8 by 8 pixels and uses few microW of power per pixel. (full report)

Electronic Nose Project
Samuel Tang, Rodney Goodman

The proposed electronic nose chip is composed of four parts: sensor stage, signal processing stage, database, and classifier. (full report)

VLSI Implementation of a Neural Network
Vincent Koosh, Rodney Goodman

We are developing a single chip solution to implement a feedforward neural network and training algorithm. (full report)

Microbat
Nick Pornsin-Sirirak, Yu-Chong Tai
Collaborators: Hany Nassef (UCLA), Chih-Ming Ho (UCLA), Joel Grasmeyer (AeroVironment), Matt Keennon (AeroVironment)

Through the discovery of flapping-wing (unsteady-state) aerodynamics, the world's first electric-powered palm-sized ornithopter has been successfully developed and test-flown. This effort is enabled by the use of a new titanium-alloy MEMS (Micro-Electro-Mechanical Systems) airframe/wing technology to produce light but robust 3-D wings. Parylene-C is used as wing membrane. This new wing design results in a 40% wing area reduction compared to the 1st generation wing. We have built a system that includes a lightweight NiCd battery and an electrical motor, a gearbox transmission design of 22:1 gear ratio with 90% efficiency, and a DC-to-DC voltage converter. Together, it allows us to design a complete system with necessary components within the weight budget for a successful flight. So far, the best flight duration obtained by Microbat was 18 seconds. It is mainly limited by the power source. (full report)

Micromachined Fluidic Couplers
Ellis Meng, Shuyun Wu, and Yu-Chong Tai

Several types of silicon fluidic couplers have been designed, fabricated, and tested for the purpose of facilitating external connections to MEMS fluidic devices. By using both bulk micromachining and DRIE techniques, couplers of various geometries have been produced for use with any standard MEMS fluidic port. Furthermore, couplers exhibit excellent performance, having an operating range of at least 0-1300 psi. (full report)

Polymer Based Electrospray Chips for Mass Spectrometry
Xuan-Qi Wang, Amish Desai, and Yu-Chong Tai
Collaborators: Lawrence Licklider, Terry D. Lee (Beckman Research Institute, City of Hope Research Center, Duarte, CA)

We have developed a MEMS system with an overhanging polymer microcapillary 2.5 mm in length and with a 5 µm x 10 µm orifice size at the tip. The fabricated systems have been successfully interfaced with a mass spectrometer (MS) to validate electrospray ionization (ESI) for biochemical analysis. The prediction of a reduction in Taylor cone size has also been observed with real time ESI fluid visualization from our chip. Built-in micro particle filters and centimeter long serpentine microchannels were fabricated on the chip with a low temperature process by using the Parylene polymer as a structural material, aluminum and photoresist as sacrificial layers, and bromine triflouride (BrF3) gas phase etching for final microcapillary releasing. The use of an overhanging polymer structure adds a new a level of mechanical robustness that was never achievable with other thin films. Functionality of our device was proven by consistent detection of Myoglobin in a 200 nM solution at a flow rate of 35nL/min and a voltage potential of 1.5 kV. (full report)

MEMS Flow Sensors for Nano-Fluidic Applications
Shuyun Wu, Qiao Lin, Yin Yuen, and Yu-Chong Tai

We have developed micromachined thermal sensors for measuring liquid flow rates in the nanoliter-per-minute range. The sensors use a boron-doped polysilicon thin-film heater that is embedded in the silicon nitride wall of a microchannel. The boron doping is chosen to increase the heater's temperature coefficient of resistance within tolerable noise limits, and the microchannel is suspended from the substrate to improve thermal isolation. The sensors have demonstrated a flow rate resolution better than 1 nL/min, as well as the capability for detecting micro bubbles in the liquid. Heat transfer simulation has also been performed to explain the sensor operation and yielded good agreement with experimental data. (full report)

Micromachined Rubber O-ring Micro-Fluidic Couplers
Tze-Jung Yao, Yu-Chong Tai

The goal of this project is to develop a "quick-connect" for microfluidic devices. We have developed a simple silicone-rubber O-ring MEMS coupler. The MEMS O-ring couplers are easy to fabricate and use, reusable, can withstand high pressure (>60psi), and provide good seals. To demonstrate this concept, a quick-connect coupler between a glass capillary tube and a silicon chip has been fabricated and tested. More than 60 psi seal has been achieved between a glass tube (860 µm O.D.) and a rubber O-ring (400µm I.D.) without measurable leakage. (full report)

Super Manueverable UAV Controlled by M³ System
Fukang Jiang, Charles Grosjean, Yong Xu, Yu-Chong Tai
Collaborators: Chih-Ming Ho (MAE, UCLA), Ray Morgan, Martyn Cowley, Scott Newbert (AeroVironment Inc.)

An aircraft for the future - having no tail, controlled by M³ systems, and with no traditional control surfaces - will be developed for low altitude surveillance. A new robust system of distributed microsensors and microactuators, with associated microelectronics (a M³ system) will be designed and fabricated to satisfy flight test requirements. A new aircraft will be designed from scratch to accentuate the concept of achieving aerodynamic maneuvering through a micromachine-based deformable smart surface. This new aircraft design concept can significantly reduce weight, overall power consumption and radar cross-section. (full report)

Distributed Turbulent Flow Control by Neural-Networked MEMS
Zhigang Han, Qiao Lin, Xuan-Qi Wang, Fukang Jiang, Thomas Tsao, Yu-Chong Tai
Collaborators: Vincent Koosh (Caltech), Rodney Goodman (Caltech), James Lew (MAE, UCLA) , Chih-Ming Ho (MAE, UCLA)

The ultimate goal of this project is to develop fully integrated MEMS with microsensors, microactuators, and microelectronics (M³) for turbulent boundary layer control. We have developed many generations of MEMS shear-stress sensors for vortex detection. The latest one is a fully integrated shear-stress sensor using a post-IC process that is added onto foundry-processed CMOS wafers. This shear-stress sensor uses a gate-polysilicon hot-wire as the sensing element that sits on a freestanding Parylene diaphragm suspended over a cavity. A special Parylene vacuum sealing and etch back process is used to achieve better thermal isolation and overall sensitivity. Wind tunnel testing of this sensor shows a sensitivity of 30 mV/Pa and a measured bandwidth of 18 kHz. We have also performed extensive theoretical analysis of these sensors. The resulting 2D MEMS shear-stress sensor theory, which includes heat transfer effects ignored by the classical theory, is verified by experimental data. We also perform 3-D heat transfer simulation and the results agree with the testing data and support the proposed new theory. (full report)

Actuated Surgical Endoscopes for Minimally Invasive Surgery
Hans D. Hoeg, Joel W. Burdick, A. B. Slatkin

Our effort is aimed at developing articulated surgical endoscopes that can access the interior of the human body in a minimally invasive manner for the purposes of visualization, diagnosis and therapeutic intervention. We have specifically focused on design and construction of scopes for use in brain surgery and gastrointestinal procedures. (full report)

The Bin Model for Generalization
Alexander Nicholson, Xubo Song, Yaser Abu-Mostafa

The problem of overfitting the data is attacked by using the Bin Model analysis. This provides a method of bounding generalization error without sacrificing valuable training data. (full report)