Micro Air Vehicles - Special Challenges

Micro Air Vehicles - Special Challenges

Jim McMichaelGeorgia Tech Research Institute

Presented to the DARPA Airplane on a Chip Workshop

Arlington, VA6/20/06

Micro Air Vehicles

Original Technical Objectives: Develop Flight Enabling Technologies Develop and Demonstrate Micro Air Vehicles Capable of

Sustained Flight and Useful Military Missions

MAV Definition: Small Air Vehicle No Larger Than

15 cm. in any dimension.

Fully Functional Vehicle Capable of Performing a Useful Military Mission at an Affordable Cost.

MAVS: A Prophecy Yet to be Fulfilled!

Why Micro? Local, On-Demand Situational Awareness Organic Asset, Eliminates Latency

Enables Completely New Missions Urban canyons, building interiors, ...

Vehicle weight - Trades with Water, Ammo, ... Eliminates Logistics Tail Affordable (Even Attritable) Hard to Detect Simple, Easy to Operate A DARPA-Hard Problem

MAV Provides “Over-the-Hill” Reconnaissance

• 15 cm, Fully functional military air vehicles

• Local situational awareness for small units

• Platoon level asset• Eliminates latency

• 30-60 minutes, 3-10 km• Day/night imaging

SimplicityLow cost

Soldier Proof

MAVs for Urban Operations

MAV HOST

MAV’s can deliverunattended

surface sensorsInterior Operationand Surveillance?Interior Operation?

Sensor Placement, Perching

Reconnaissance

See thru Windows

Bio-Chemical SensingChemical CloudTracked by MAV

Sensor detectsPPM - PPB

MAV Assisted Pilot RescueMAV provides situational awareness,

Provides beacon for rescue operations.

High Level Challenges• Low Re Aerodynamics and Control

• 15 cm MAV Re ~ 100,000

• Lightweight Power and Propulsion

• Navigation, guidance and control, autonomy, &communications

• Ultra-light structure, sensors and payloads

Speed

Hover

Agility

Covertness

High Degree of Integration and Multifunctionality

Range

State of the Art (‘96)Model Aircraft (>30 cm)• High Re Physics• High aspect ratio, high wing loading• Man-in-the-loop Radio Control

Model Airplane Engines ( .01 cu in)NiCd and Li Batteries• 300 mW/g• 350 J/g

Radio Control• GPS - (14 g, 3 W, )• Gyro - ( 28 g, .1 W, 1°/hr drift)• Comms. - (7 g, 1 W, 10 km)

Low Resolution Video • (380x540, 15g, 2watt)

MAV Technical Challenges

GoalsFixed Wing, Hover, Flapping (<15 cm)• Low Re Aero Performance• Low aspect ratio, low wing loading• Autonomous flight controls

Very Small ICEs (0.0035 cu in)Fuel Cells, Micro Turbine Engines

• 500 mw/g• 3,000 J/g, (30 g max)

Autopilot (15 g, 1.0 W)• GPS - (3 g, 0.2 W, )• Gyro - ( 3 g, .1 W, 0.1°/hr drift)• Comms. - (2 g, .5 W, 4Mb/s)

High Resolution, Low light• (1000x1000, 4:1 Compr., 3g, 0.2W),• IR Imager (<15g, <1W)

Enabling Technologies1. Aerodynamics, Stability

and Control

2. Ultra Lightweight Propulsion and Power

3. Guidance, Nav, and Comms

4. Sensors and Payloads

The Essence of the MAV Challenge

• Reduce Weight, volume, power required

• Increase complexity, reliability, functionality, utility ....... That can be put into a small, highly integrated package

“Original” (Natural) MAV Program Solved the Complexity Problem

Airframe & Materials

Unique Missions: Perch and Stare

Aerodynamics & Propulsion

Integration &Multifunctionality

10 g.10 cm wingspan3 min.Acoustic sensorsMEMS wings

Micro Air Vehicles

Caltech AeroVironment

Black WidowMicrobatLutronix

KolibriMicroSTAR

Lockheed Sanders

Stanford Research Institute

100 g.5 km range, 30 min.Autonomous Nav.Video imagery,Vmax ~40 m/sAvionics Pow ~ 2 WPayload ~ 4 g, 6 cm3

Mentor

320 g.30 min.Hover/translateGPSAutopilotVmax ~ 10 m/s Avionics Pow ~ 4 WPayload ~ 15 g, 95 cm3

50 g.Electrostrictive Polymer Artificial MuscleFlapping flightAvionics & Payload ~ 5 g

50 g.1 km rangeTeleoperatedVideo imageryVmax ~ 20 m/sAvionics Pow ~ 2WPayload ~ 2 g, 2 cm3

g % g % g %

Avionics 78 25 13.5 16 13 21

Power Systems 169 53 58 69 38 61

Airframe 54 17 9 11 9 14

Payload 15 5 4 4 2 4

Total 316 100 84.5 100 62 100

MAV Subsystem WeightSubsystem Weight and Mass Fraction

Power for MAVs

• Typical power for fixed wing 15 cm MAV (100g) is about 4 W for flight and 1-2W for avionics and comms

MAV Propulsive PowerComponent Size, Weight, and Power Requirements Must be Minimized

Maximize Endurance Parameter• Maximize Aerodynamic Perfromance• Thin cambered airfoils (low Re)

Minimize Altitude

Maximize Propeller Efficiency• Optimized size, speed & type may

not suit operational needs

Minimize Wing Loading• minimize wing loading• large wing area

- low aspect ratio

Minimize Weight• component synergy• microfabrication• reduce fuel load• maximize propulsion system energy density

• multifunctional materialsand components

CD3/2

WS

2ρPower to fly = W η

1/21/2

CL

MAV Propulsion and Power Options

TechsburgThermoelectrics

Projected 400 mw/g

IGR Solid Oxide Fuel Cell

M-DOTMini Turbine Engine

18 Watts 42 grams 4 hours 6200 J/g 430 mW/g

Aerodyne MICE

10 Watts 21 grams 1200 J/g 330 mW/g

( for 1 hour endurnace)

D-STARMini Diesel Engine20 Watts, 80 Watts

World’s smallest gas turbine1.5x3.25 in78 gm1.4 pounds of thrust1.27 KWJP8 - diesel

H2 Demonstration Micro-Engine

Thrust = 11 g

Fuel burn = 16 g/hr

Engine weight = 1 gram

Turbine inlet temp = 1600°K (2421°F)

Rotor speed = 1.2 x 106 RPM

Exhaust gas temp = 970°C

Silicon-Cooled Turbine

Need “Precision Microstructures”

Estimated Power Subsystem Options for 100 g Fixed Wing MAV

• Power System Requirements for 100g Fixed Wing MAV– Power system weight allocation is 60% vehicle weight, 60 g– 4 W peak available propulsive power, 3.3 W average– 5 W continuous conditioned electric power

• Endurance in hours for different power system combinations

Power System Combinations

Hou

rs

5432

1

SOFC/motor

5.4

MicroTurbojet/Turbogenerator

1.2

MicroTurbo-generator/motor

4.5

Triple ICE/Alternator

1.5

Diesel/Alternator

3.2

Cox-.01

0.6

E-PicherBattery

& Motor

0.2

MAV Payload VolumeCurrent MAV Systems Estimated Payload Bay Dimensions

Lutronix

1.9 cm height, 4.8 cm diameterVol = 95 cm3

Sanders

5 cm length, 1.1 cm height, widthVol = 6 cm3

1.2 cm cube, Vol = 2 cm3

AeroVironment

Need Increased Volume for Payload Sensors or Smaller, More Integrated, More Capable Avionics, Sensors and Processing

Data Link

Processor

Imager

Gyros

Rudder

Accelerometers

Engine

Con-troller

FlapActuators

FlapActuators

Nav/Positioning5 Grams

160 mW

3-AxisMagnetic

Pressure Sensor

Processor/Memory2.5 Grams600 mW

Camera/Lens4 Grams350 mW

Datalink/Antenna6.5 Grams2,000 mW

Airframe7 Grams

Batteries44.5 Grams

13 Watts AvailableAntenna

Power System

Actuators2 Grams200 mW

Engine/Prop13.5 Grams7,000 mW

Total System85 Grams

13 W

MicroSTAR Conceptual Layout (1996)

“Little big-airplane design paradigm” is followed in essentially all present MAV designs (2006)

1996 Autonomous Flight Control System Conceptual Layout for Kolibri

Avionics Assembly

3-Axis MCM-L Accelerometer

3-Axis MCM-LGyroscope

RF Communications/Proximity Sensor Electronics

GPS Module

Flex Backplane

Flight Control Computer

Battery

Power ModuleStabilator Cntrl

Image Sensor

Altimeter

Avionics FrameAcc

els

Gyr

osC

omm

s /P

roxi

mity

/ R

FG

PS /

FCC

Bat

tery

Pow

er

Avionics Top View

20 min endurance10-60 mphTurn radius = 15 ft.Span = 20cm

“With current COTS electronics, it is very difficult to carry any useful sensors or flight control system in a 14 cm MAV and still achieve decent duration.”

“When the Ground Control Station is added, the overall system sizes for a 15 cm and 60 cm MAV are about the same.”

MLB’s Trochoid MAV

Without increased multifunctionality and integrated components, the potential advantages of MAVs are not realized.

Black Widow1

AerovironmentWASP2 (41 cm) Aerovironment

University of Arizona4 University

of Florida6

University of Notre Dame5

MicroStar3

BAE Systems

Sample of Current “MAVs”

0

10

20

30

40

50

60

70

80

90

100

0 50 100 150 200 250 300

Span (cm)

ND

Pointer

Dragon Eye

OAVCyber Bug

MicroSTAR

WASP

Black Widow

Endurance Falls Rapidly as MAV Scale Decreases

Endurance Rises Rapidly as Weight Increases

Is There A “Size Asymptote for Utility” Without Higher Levels of Integrated

Multifunctionality?

20 cm Trochoid

MLB Company

OAV9

Honeywell

The “DARPA - Easy”* Problem

* “If you want more capability:Grow it bigger”

The DARPA-Hard problem is still to increase functionality at reduced scales through multifunctional integration

QuickTime™ and aVideo decompressor

are needed to see this picture.

‘Typical’ Rotary Wing MAV in Flight

QuickTime™ and aVideo decompressor

are needed to see this picture.

The views, opinions, and/or finding contained in this article are those of the author and should not be interpreted as representing the official policies, either expressed or implied, of the Defense Advanced Research Projects Agency or the Department of Defense

What Might Be Possible With Much Higher Levels of Integration & Embedded Functionality?

• Highly capable MAVs as a testbed for autonomous on-board decision making, both tactical and strategic– Develop revolutionary operator interface paradigms

• Intelligent, collaborative, self aware systems (air and ground sensors and vehicles)

• Real time SA advisor for operators in the field.

Return to the DARPA-Hard Challenge:- Truly Mission Capable MAVs

MEMS and Airplane on Chips

Amit LalProgram Manager, DARPA/MTO

Airplane-on-a-Chip Workshop, June 20-21, 2006

Outline

• Background on MEMS• AOC and MEMS• Insect on Airplane on Chip

MEMS?

• Micro • Electro• Mechanical• Systems

Lithography based , and perhaps self-assembled –not hand assembled

Electrical engineering –Integrated circuits, devices, circuits, multiphysics

Mechanics and fluidics of IC-fabricated structures –integration of mechanical with electrical –electromechanics

If only mechanical – the field is micromechanics

Systems: Means many thing to many people – the least well addressed vision of MEMS – but DARPA is at the forefront

Progression of MEMS

1950 1960 1970 1980 1990 2000

HNAAnodic Bonding

EDP

Pressure Sensor (Honeywell)

KOH

Si Pressure Sensor(Motorola)

Si as a mechanical material (Petersen)

SFB

SFB Pressure Sensor (NovaSensor)

TMAH

DRIE !!

XeF2/BrF3

Metal sacrificial process (US Patent)

RGT (Nathanson et al)

Metal Light Valve (RCA)

PolySi beams (Howe, Muller)

PolySi Micromotor (Tai, Muller) IR imager (Honeywell)

PolySi Comb Drive (Tang, Howe)

ADXL AccelerometerDMD (TI)

Si Gyro (Draper)LIGA

IC

RF MEMSBio MEMS

CSAC, NGIMG

Fabrication plus basic sensors

Surface micromachined sensors

2010

MicroElectroMechanicalSystems

CMOS microelectronics,

RF

III-V compound

semiconductor

Nanotubes, QDs

RadioactivityGas/vaccum

devices

Nanofabrication

CryoelectronicsMicrofluidics MEMS

MEMS Electrostatic Actuators

Suspended structure

Example: MEMS Automotive Sensors

G= Gyro

A= Acceleration

•L=Low (<5g)

•M=Medium (50g)

•H=High (>100g)

Airbag DeviceNon-Airbag Device

AH (Z or X)

airbag

airbag

airbag

alarmAL (XY)

airbagGY

GZ

GZ

AL (Z)

rolloverAL (ZY)

VDCAL (XY)

GPS/NavAL (XY)

AM (Z or X)

AM (Z or X)

AM (XY)

•Small proof mass enables fast response enabling side impact bags•Small size enables low power operation•Small size enables high thermal isolation and budgets•Small size enables high resonant frequencies

MEMS and Transistors

Historic MEMS Devices

Future MEMS Integration Levels Enabled Applications

Increasing Ability to Sense and Act

Incr

easi

ng A

bilit

y to

Com

pute

100

101

102

103

104

105

106

107

108

109

101 102 103 104 105 106 107 108 109

Number of Mechanical Components

Optical Switches and Aligners

Parts Handling

Integrated Fluidic Systems

Physiological Sensors

Inertial Navigation on a Chip

Terabits/cm2

Data Storage

Ultrasonic Imagers

Displays

RF Switching, Wireless

Distributed Structural Control

Num

ber

of T

rans

isto

rs

Cars, Consumers

Micromachining: Users viewpoint

• Technology to make microscale mechanical parts integrated with electronics

• Another machining technology, like mill and lathe• Mill and lathe were really exciting when they first

came out - now we take them for granted• DARPA MEMS Exchange is a program for users!• Various MEMS foundries also serve this purpose

Micromachining: Integrated Circuits Viewpoint

• Micromachining is a bag-of-tricks on top of the conventional IC-fabrication facility

• Lithography is used to generate microscale mechanical features

• Development in finer resolution lithography means smaller mechanical machines (e.g. DUV lithography)

• MEMS is usually “dirty” for ultraclean circuit fabs

MEMS is like Spanish moss on

the IC industry tree

The New Paradigm Viewpoint

• Micromachining enables a new way of thinking about mechanical structures

• Parallel (huge) arrays of machines are possible

• Machines comparable to size of biological objects are possible

• Integration of electromechanical machines to individual small scale objects is possible

AOC and Selected MTO MEMS

Cs or Cs or RbRbGlassGlassDetectorDetector

VCSELVCSEL

SubstrateSubstrate

GHzGHzResonatorResonatorin Vacuumin Vacuum

PhotoDetector

VolVol: 1 cm: 1 cm33

Power: 30 Power: 30 mWmWStab: Stab: 11××1010––1111

Chip-ScaleAtomic ClockCSAC

Atomic Sensors

Flight Control

NGIMGZ

YX

HERMIT

Micro Flaps

3D MERFS

ASP

MIT μturbine

μfluidic fuel injector

μcombustion

µWankelMCC

MGA

Atomic Clock Principle & Design

• Problem: µwave cavity dictates size133Cs

m = 0f = 4

m = 0f = 3

m = 1

852.11 nm

852.11 nm

ν = ∆E/ħ = 9 192 631 770 Hz

9 192 631 770 Hz

133Cs vapor at 10–7 torr852.11 nm laser

Photodetector

µ-wave cavity

Microwave at9 192 631 770 Hz

Optical-Microwave Atomic Clock

µ-wave f

852.11 nm laserModulated at

9 192 631 770 HzPhoto

detector133Cs vapor at 10–7 torr

Mod f

All-Optical Atomic Clock• Solution: optical excitation (CPT)

• Frequency determined by an atomic transition energy

InterrogatingLaser

Absorption

HyperfineTransition

Input Energy

Chip-Scale Atomic Clock (CSAC)PM: Amit Lal

Goal:• Integrate MEMS, photonic, and electronics

technologies to achieve miniature, low-power atomic timing and frequency references with

– Allan deviation < 10-11 over 1 hour (1µs/day)– Size <1 cc– Power Consumption < 30 mW

Technical Challenges:• Cell design for maximum Q• GHz high-Q reference resonator• MEMS-enabled thermal isolation for low powerKey Accomplishments:• Chip-scale atomic physics package demonstrated

with ~5 mW of power consumption• GHz vibrating micromechanical resonators

demonstrated• Chip-scale atomic clock with 5x10-11 Allan deviation

at 1 sec, in a 10cc package, consuming 180 mWImpact:• High-security communications• High-confidence ID of friends and foes• Ultra-sensitive radar• Missile and munitions guidance• Longer autonomy (radio silence interval)• Tech Transfer Path: Air Force, JTRS

communications, Army, numerous others in GPS

3.53

cm

Physics Package

3.94 cm Physics Package

Precision Time for Every Radio and Network Node

0.0001 0.001 0.01 0.1 1.0 10.0 100.0 1000.0

0.0001

0.001

0.01

0.1

1.0

Bias Drift [o/hr]

Rate Grade

Tactical Grade

Navigation Grade

Ang

le R

ando

m W

alk

[o/√

hr]

Gyroscope Requirements

Rate Grade

Analog ADXRS401ARW: 3.8 o/√hr

Bias Stability: 50 o/hrFull Scale: ± 75 o/s

Power: 30 mWSize: 0.15 cc

Honeywell HG1900ARW: 0.05 o/√hr

Bias Stability: 5-12 o/hrFull Scale: ± 1,000 o/s

Power: 1.95 WSize: 283 cc

Litton HRGARW: 0.00006 o/√hr

Bias Stability: 0.0003 o/hrFull Scale: ± 4 o/s

Power: 22 WSize: 8,118 cc

Nuclear Magnetic Res. Gyroscope

• The ultimate in miniaturized spinning gyroscopes?– from CSAC, we may now have the technology to do this

Atoms AlignedNuclear Spins

0º

-20º

20º

0º

-20º

20º

LaserPolarizer

Rb/Xe Cell

Photodiode

3.2 mm

1 mm

1 mmtθ

Soln: Spin polarize Xe129 nuclei by first polarizing e- of Rb87 (a la

CSAC), then allowing spin exchange

Better if this is a noble gas nucleus (rather than e-), since nuclei are

heavier less susceptible to B field

Challenge: suppressing the effects of B field

ARW ~ 0.0006 o/√hr

Nav-Grade Integrated Micro Gyroscopes (NGIMG)

PM: Amit LalGoal:• Attain tiny gyros and accelerometers with

navigation-grade performance and tiny power consumption

Technical Challenges:• Reducing magnetic field sensitivity for NMR

gyros• Attaining high spin rate for levitated gyros• High Q optical resonatorsKey Accomplishments:• Ultrahigh 106 Q quartz resonators

demonstrated for class 2 gyro• 50 rpm rotating and electrostatically levitated

spinning mass demonstrated• Nav-grade performance in NMR gyro on an

optical bench demonstrated using B-field nulling

Impact:• Enable wearable, low power IMU modules for

dismounted soldier• Guidance for munitions, UAVs, and insect-like

robotics platforms• Tech Transfer Path: Air Force, JTRS

communications, Army, numerous others in GPS

X

Z

YRotor 1

Rotor 2

RPM ~133

Harsh Environment Robust Micromechanical Technologies (HERMIT)

PM: Amit Lal

Goal:• Attain superior performance and lifetime of MEMS devices

via localized control of their operating environments• Achieve upwards of a 10x improvement in cycle lifetime for

RF-MEMS switches, and temperature stability for resonators & gyros

Technical Challenges:• Wafer-level vacuum packaging of micro-scale devices• Chip-scale control of environment, including: pressure,

temperature, vibration, and gaseous speciesKey Accomplishments:• Wafer-level packaged RF MEMS switch demo’d with 0.02dB

loss @ 35GHz and > 1 billions cycles• CVD-based vacuum packages demonstrated with no

observable leakage over 6 months• Packaged resonator freq. drifts <2 ppm over 6 months• Much higher (95%) yields observed for wafer-level

packaged micromechanical resonatorsImpact:• Enable small, low-power, true-time delay units for phased

array antennas, reconfigurable filters, and low-phase noise oscillators for radar and communications

• Enable guided munitions and self-navigating micro-air vehicles

• Tech Transfer Path: Army, Air Force, DoD T&E

Realizing the Promise of MEMS through Wafer Level Packaging

Robust Capacitive MEM Switches

CVD CageStructure

OpenVA

Tri-StateLogic

Seal via combination of spin-on-glass

(SOG) and nitride

Seal via combination of spin-on-glass

(SOG) and nitride

Package Insertion Loss - Cages + SOG

-0.5

-0.45

-0.4

-0.35

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

0 5 10 15 20 25 30 35 40 45 50Frequency (GHz)

Inse

rtio

n Lo

ss (

dB)

Thru1Thru2514B1cs514B2cs514B3cs514S1cs514S2cs514T1cs

0 10 20 30 40 50Frequency [GHz]

Inse

rtio

n Lo

ss [d

B]

0

-0.1

-0.2

-0.3

-0.4

-0.5

Thru Line

Package (Cage + Encapsulant)

The HERMIT Concept and Impact

Gap ~120 nm

UNCD

Stanford – Embedded resonators - timing

Michigan - Gyroscopes

UCB–SiC Strain Sensors, T up to 600°C

Honeywell – Liquid Metal, High Power

Northeastern – High Force Ohmic

Memtronics – Capacitive

UCSB – Ti-MEMS

Argonne – CMOS Compatible Diamond MEMS

Resonators/Sensors

Switches

Radant

Micro Gas Analyzers (MGA)PM: Dennis Polla

Goal:• Enable remote detection of chemical agents via

tiny, ultra-low power, fast, high sensitivity, chip-scale gas analyzers with low incidence of false positives..

Technical Challenges:• Rapid chemical separations (GC).• New detectors (mass spectrometers, chemi-

resistors, Nanotube detectors).• Sensitivity < 50 ppt• False alarm rate < 1/200,000 measurements.

Key Accomplishments:• Preconcentrations of DMMP >8,000• Separation of 4 CW simulants in the presence of

4 hyrdocarbon interferrents• 1 Torr operation of a 1-mm ion trap mass

spectrometer with 2.3 amu resolutionImpact:

• Deployment of highly sensitive analytical instruments in the field with a low incidence of false-positive detections

PO

O OCF3

CF3

OH

Si

Si

O

Si

O

SiO

Si

O

Si

O

Si

OO

Si

O

Si

Si

O

OSi Si

OSi

OSi

OSi

O

SiO

O

Si

O

Si

O

Si O

SiO

O CF3

F3COCF3

F3C

O

O

OP

H

OO

OP

H

Hydrophobic CF3 groups

Strong hydrogenbonding interactions

Absorbent covalentlybonded to silica matrix

Micromachinedchromatography

columns

Chemical detectors

Ion Mobility Mass Spec

Chemicalpreconcentration

micro/nanostructures

Chemicalfunctionalization ofsorptive molecules

Gold Standard Chemical Gas Analyzer in a Match Box

Micro Cryogenic Cooler (MCC)PM: Dennis Polla

Goal:• Attain superior performance (e.g. , of LNA, RF front

ends, sensors, …) by cooling to cryogenic temperatures

• Power consumption <200mW• DTBC in isolation with >10x perf. increase• Ex: COP: >0.01 @77K, >0.04 @160K• Size < 4cc

Technical Challenges:• Active or passive circuits on micromachined

platforms• Micromachined compressors, heat exchangers, high

pressure• Thermoelectric cooling, laser cooling, Joule-

ThompsonKey Accomplishments:• Proven gettered vacuum packaging methodology

directed self-assembled wafer-level fab allows the use of an unconstrained variety of TE’s

• Tiny diameter photonic crystal heat exchanger tubes→25atm

• Heat exchangers supported by thermally isolating SU-8 rods

• Piezoelectric compressors → 25:1 pressure dropImpact:• Imaging at terahertz frequencies (defined roughly as

300GHz-3THz)• Increase application performance of IR detectors,

low noise amplifiers and data converters, electromechanical oscillators, and high temperature superconductor circuits

3.94 cm

Images of 3-D MERFS Structures (5 lithographic layers, 3 Material)

PM: John Evans

1 mm1 mm

Dielectric Strap

Center copper conductor

Outer copper conductor

Hybrid Coupler

Isolation test structures

Attenuation Test Structure Cavity Resonator

“Launch” De-imbedding Test Structures 6-Inch 3-D MERFS Wafer on RF Probe Station

λ/4 matched flexure tether

Drive and Sense electrodes

30 nmHafnium Dioxide

Silicon

λ/4 matched flexure tether

Drive and Sense electrodes

30 nmHafnium Dioxide

Silicon

ASPPM: John Evans

ADCIF Chain

IF Chain

IF Chain

IF Chain

IF Chain

RF FilterBank20 MHz –

6 GHz

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

25 MHz BW

70 MHz IF

20 MHz –6 GHz

Analog Sensor

5 GHz/s requires 5 ms per 25 MHz block.

5 ms / 100 us sample periods

50 sample periods per 25 MHz block

IF Chain

ADC

ADC

ADC

ADC

IF Filter

IF Filter

IF Filter

IF Filter

IF Filter

Distribution Statement “A” (Approved for Public Release, Distribution Unlimited), DARPA Case 6709, 2/15/06

MEMS wing boundary layer modification

Typical flow over a delta wing at moderate angles of attack

MEMS shear stress sensonr and shear stress “imager”

Ho et al. (UCLA)

Micromachined sensors and actuators could

provide sufficient moments/torques to replace large flaps in aerodynamic control

Mico flap actuator under activation

1st generation flexible shear stress sensor

2nd generation flexible shear stress sensor

Small disturbances caused by microactuators at appropriate locations could cause appreciable changes in the

global flow field

Insect-on-Chip

WASP-DARPA, 200 grams, 12-inch wing span, 1-2 hr fly time

Caltech flyer

UC Berkeley Flyer

Nano - Air Vehicle Program

•20-30 minute mission•7-cm wingspan•Controlled hover and landing

Insect locomotion is used as model for these efforts

Why not use insects directly?Bee training, insect backpacks -> Insects too unpredictable (temperature, wind, humidity,

mating, feeding, etc.)Insight: Eliminate unnecessary biology using

MEMS

Insect Cyborgs

Normal growth

DARPA Program : Use object

insertion ability into pupas to reliably

insert microsystems

(instead of glass tube) for insect

control

Pupa halved and front develops into

moth

Sectioned Pupa with pipe inserted for hormone transport – grows into moth shown above. Insertion of

chemical blocking ball bearing results in no growth

Platform with Silicon Chips

Balsa platforms were stitched in the pupae stage

Moth emerges with platform and flies for 2 weeks

•We have proven successful platform insertion in pupa stage, without effecting flight, or lifetime•Ready for locomotion control payloads!

Hybrid-INSECT MEMS

Ultrasonic transducers

Pheromone ejectors

Light sources

Piezoelectric flaps for power

Thermoelectric power, flexible

platform

Cross and across inserts in pupae

Platform for sensors, actuators, and comm

NGIMG

CSAC

•Moth body weight ~1-5 grams•Payload ~ 0.5-5 grams

Microfluidic port

Tissue-anchors

Neural/Muscleprobes

Summary

• MEMS offers pathways to miniaturized and chip-scale sensor and actuator systems to reduce size of AOC components

• Upcoming MEMS/CMOS integrated solutions will enable cost reduction of AOC systems

• Integration of MEMS with nature (e.g. insects) could provide methods to realize cyborgs that might require unique AOC capabilities

QUESTIONS?

Amit Lal, DARPA-MTO1

Microsystems, Scaling, and Integration

Amit Lal, Program ManagerMTO/DARPA

Microsystems Technology SymposiumSan Jose, CA, March 6, 2007

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE 06 MAR 2007

2. REPORT TYPE N/A

3. DATES COVERED -

4. TITLE AND SUBTITLE Microsystems, Scaling, and Integration

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) DARPA

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited

13. SUPPLEMENTARY NOTES DARPA Microsystems Technology Symposium held in San Jose, California on March 5-7, 2007.Presentations, The original document contains color images.

14. ABSTRACT

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT

UU

18. NUMBEROF PAGES

21

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18


Progression of MEMS

NEMS-CMOS: Meshing of transistors and relays

Hybrid-Insect MEMS

Integrated RF-SYSTEMS - RADAR

Chips-Scale Atomic Clock

1950 1960 1970 1980 1990 2000

HNAAnodic Bonding

EDP

Pressure Sensor (Honeywell)

KOH

Si Pressure Sensor(Motorola)

Si as a mechanical material (Petersen)

SFB

SFB Pressure Sensor (NovaSensor)

TMAH

DRIE !!

XeF2/BrF3

Metal sacrificial process (US Patent)

RGT (Nathanson et al)

Metal Light Valve (RCA)

PolySi beams (Howe, Muller)

PolySi Micromotor(Tai, Muller) IR imager (Honeywell)

PolySi Comb Drive (Tang, Howe)

ADXL AccelerometerDMD (TI)

Si Gyro (Draper)LIGA

IC

RF MEMSBio MEMS

2010

MicroElectroMechanicalSystems


Two views of MEMS

MEMS is like Spanish moss on the IC industry

tree

http://www.mems-exchange.org

MEMS for everyone/everything?


MEMS for Microsystems

• Miniaturization/Integration – SWAP• Scaling for higher performance• Multiphysics• Biological interfaces• Gateways to nanoscale effects• Environmental control over sensors and

actuators

Temex RMOVol: 230 cm3

Power: 10 WAcc: 1×10–11

Symmetricom CSACVol: 7.8 cm3

Power: 95 mWStab: 5×10–11/100s

100 µm

Drain Source

Gate

Contact Detail

Beam

RF-MEMS switch

Integration of Alkali-metal

vapor on chip for atomic sensors

Embedded MEMS - HERMIT

Universal MEMS package-HERMIT

0.8 cm

Navigation grade Gyroscope

Insect MEMS

CSAC

NEMS - switch


Radant Demonstrates>900 Billion Switch Cycles

Wins Frost & Sullivan Excellence in Technology Award

Tri-Service DoDTesting Team

MEMS:Undeniable Reliability

PM: Amit Lal, HERMIT

Demo Radar

0.4 m2 Azimuth Scanning MEMS RadantTM Lens

Composite Frame (Graphite / Epoxy)

APG-67Xmtr

Lockheed Martin Modified APG-67 Radar Components APG-67

ProcessorFeed

ControlLens Ctrl/Interface

APG-67RF

Ctrl / Interface

New / Modified HW/SW

Modified Hardware

30 degree scan 0.4m2 ESA

MEMS Insertion into the RadantTM

Lens Architecture has Been Demonstrated

This Antenna is the First Large Scale Use of MEMS Switches in the World

MEMS Insertion into the RadantTM

Lens Architecture has Been Demonstrated

This Antenna is the First Large Scale Use of MEMS Switches in the World


Hybrid-Insect MEMSVISIONVISION

OBJECTIVESOBJECTIVES• Develop technology to enable highly coupled electro

mechanical interfaces to insect anatomy • Demonstrate MEMS platforms for electronic locomotion

control, power harvesting from insect, and eliminate extraneous biological functions

Create technology to reliably integrate microsystems payloads

on insects to enable insect cyborgs


Background: Insect Metamorphosis

1st instar 2nd instar 5th instar4th instar3rd instar

Storage of energy over weeks to use later for flight


Key Experiments in 1940s

Normal growth

DARPA Program : Use object

insertion ability into pupas to reliably

insert microsystems

(instead of glass tube) for insect

control

Pupa halved and front develops into

moth

Sectioned Pupa with pipe inserted for hormone transport – grows into moth shown above. Insertion of

chemical blocking ball bearing results in no growth


MEMS PlatformUltrasonic

transducers

Pheromone ejectors

Light sources

Piezoelectric flaps for power

Thermoelectric power, flexible

platform

Cross and across inserts in pupae

Platform for sensors, actuators, and comm

NGIMG

CSAC

•Moth body weight ~1-5 grams•Payload ~ 0.5-5 grams

Microfluidicport

Tissue-anchors

Neural/Muscleprobes


HI-MEMSHybrid Insect MEMS

PM: Amit Lal

Microsystem platform inserted into moth in pupae stage, and successful emergence of adult moth

with microsystem

X-ray images of probes in muscles show good tissue

growth around inserted probes

Boyce Thompson Institute:Insect Sentinals


Hybrid NEMS ElectronicsRelay computer

(circa 1950)

Pentium (2006)

NEMS/CMOS

Abacus Babbage

4004 (1971)

8086 (1978)

+


Hybrid NEMtronicsObjectivesObjectives

• Eliminate leakage power in electronics to enable longer battery life and lower power required for computing.

• Enable high temperature computing for Carnot efficient computers and eliminate need for cooling

ApproachesApproaches• Use NEMS switches with and

without transistors to reduce leakage – Ion:Transistor, Ioff: NEMS

• NEMS can work at high temperature, enabling high efficiency power scavenging.

N+ N+ P+ P+

N-WellP-Substrate

VDDOUT

GND ININ

All Mechanical Computing

Hybrid NEMS/CMOS component integration

Hybrid NEMS/CMOS Device integration

1

1

0

0

1

0

01

Ion

Ioff


Nano Switches

Released FinFET NEMS switch

Nanotube/Fiber switches

Nano-machined switches

50 nm tines

CMOS Integrated NEMS

40 nm beam

Nanoscale e-

shuttle

1 µm in 0.35µm,

100nm in 90 nm CMOS


The Problems: Max Heat Removal Rate and Leakage Power

Lg/VDD/VT trends increases in:• Active Power Density (∝VDD

2) ~1.3X/generation• Passive Power Density (∝VDD) ~3X/generation

Excessive Heat Generation

NEMS

1E-05

1E-04

1E-03

1E-02

1E-01

1E+00

1E+01

1E+02

1E+03

0.01 0.1 1Gate Length (µm)

Pow

er (W

/cm

2 )

Passive Power Density

Active Power Density

Excessive Ioff

NEMS can eliminate

leakage current –

Zero-idle

powerNEMS can work at high

temperatures: Carnot

efficient computing


The Carnot Optimized Computer

PextSi

+

Pconv

TH

TC

)1(H

CHGinin

H

CHGinext T

TTPPT

TTPP −−=

−−= ηη

• TH should be maximized for high Carnot efficiency• 700C => 973-300/973 = 0.70 • If 50% of Carnot => 35% power can be reclaimed• Cooling could be eliminated• Needs fast switching technology at high temp – NEMS

PinPremove

Past Example


Self-calibrating Micro Sensors: Shoe-Implanted Perpetual Personal Navigation

CMOS-MEMS Micro 3-axis accelerometer/gyro possible but have offsets due to imprecise fab. Develop ppmaccurate sensor model using on-chip calibration techniques – eliminate temp control to reduce power

Sonic pulsing, fluid MEMS to sense velocity directly

Precision and stable resonators provide frequency for self-calibration

Power scavenging from motion in shoe ~ 10 milliWattaverage over mission

HI-MEMS insect power output >5 milliWatt average

1 cc, 5-mW average

IMU

State-of-Art (without electronics or GPS) IMU: 14cc, 250 mW

10x reduction in size, >100x reduction in power


MTO Mostly-silicon UAV

rece

iver

trans

mitt

er

IR se

nsor

acou

stic

sens

or

optic

al se

nsor

RF se

nsor

CB se

nsor

sens

ors

ante

nnae

sigpr

oces

s

fligh

t con

tr

iner

tial n

av

rece

iver

trans

mitt

er

rece

iver

trans

mitt

er

IR se

nsor

acou

stic

sens

or

optic

al se

nsor

RF se

nsor

CB se

nsor

IR se

nsor

acou

stic

sens

or

optic

al se

nsor

RF se

nsor

CB se

nsor

sens

ors

ante

nnae

sens

ors

ante

nnae

sigpr

oces

s

fligh

t con

tr

iner

tial n

av

sigpr

oces

s

fligh

t con

tr

iner

tial n

av

Distributed actuators to pump air + solar cells +

batteries

Control Electronics


Benefits of mostly-silicon MAV

Functionality – Entropy – Data Bandwidth

Wei

ght (

a.u.

) Power (a.u.)

Inertial sense guidance

RADAR

IR imaging

Collision avoidance

No-wiring, limited packaging

Low-power CMOS, MEMS components to reduce power for RF


CMOS microelectronics,

RF

Nanoelectronics

Quantum computation

RadioactivityGas/vacuum

devices

Avionics

Cryo-electronicsMicrofluidics MEMS


Summary• MEMS offers pathways to miniaturized

and chip-scale sensor and actuator systems for reduced SWAP and increased functionality

• Upcoming MEMS will result in cost/performance benefits by integrating functionality

• The future for MEMS-IC symbiosis is bright


QUESTIONS?

MEMS BASED BIOELECTRONIC NEUROMUSCULAR INTERFACES

FOR INSECT CYBORG FLIGHT CONTROL A. Bozkurt1, R. Gilmour2, D. Stern3, A. Lal1

1Cornell University, School of Electrical and Computer Engineering, Ithaca, NY 14853 2Cornell University, Department of Biomedical Sciences, Ithaca, NY 14853

3Boyce Thompson Institute, Ithaca, NY 14853

ABSTRACT

This paper reports the first direct control of insect flight by manipulating the wing motion via microprobes and electronics introduced through the Early Metamorphosis Insertion Technology (EMIT). EMIT is a novel hybrid biology pathway for autonomous centimeter-scale robots that forms intimate electronic-tissue interfaces by placing electronics in the pupal stage of insect metamorphosis. Our new technology may enable insect cyborgs by realizing a reliable control interface between inserted microsystems and insect physiology. The design rules on the flexibility of the inserted microsystem and the investigation towards tissue-microprobe biological and electrical compatibility are also presented.

1. INTRODUCTION

When Micro-Air-Vehicles (MAVs) or tiny fliers are considered, the power source required to fly them within the constraint of generating lift, powering flight control sensors and actuators, and collision avoidance has limited their mission time and autonomy. Researchers have greatly benefited from the study of naturally occurring fliers to design individual biomimetic structures as the parts of MAVs [1]. Although, a tremendous effort has been put forth to combine these structures as a complete MAV, there has been no successful demonstration of a system that can takeoff, maneuver and land autonomously for long periods of hours or days [2].

Another idea has been to directly use naturally designed and optimized flyers, namely, insects as MAVs [3-5]. Insects are self-powered, are cm-scaled and operate with highly efficient flight muscle actuators. Man-made electronic systems can be implanted in insects to study and control the insect flight by recording from and actuating either each or combinations of the sensory, neural or muscular systems. However, it is a challenge to implant electronic systems to modulate the insect’s flight without disturbing the insect’s own efficient flight mechanism. Any artificially attached platform and surgery on the adult insect is not reliable, as the inserted devices on this stage can shift, create mass-balance disturbance and cause performance affecting tissue damage, especially when the inserted electronic systems are rigid.

In our previous work [6], we demonstrated an efficient method to implant structures in tobacco hawkmoth Manduca sexta, which we call “Early Metamorphosis Insertion Technology” (EMIT). EMIT involves inserting structures to the pupae at early stages of metamorphosis (Figure 1) such that the body adapts the structures during the development and inserted structures emerge as a part of the body to create insect cyborgs. A reliable biointerface was created by taking the advantage of the rebuilding of the entire tissue system. This hybrid structure enables a platform where CMOS devices and MEMS structures can be used as sensors and actuators not only for insect flight control, but also for biological and environmental sensing. Moreover, this platform can be used to study the probe-tissue interface in general for MEMS based neuromuscular prosthetic systems.

EMIT can benefit from any insect/animal that has metamorphic development (moths, butterflies, beetles, etc.) to create insect cyborgs with different locomotion capabilities. We have selected Manduca sexta due to its relatively shorter metamorphic duration of three weeks. With its 1-2 gram carrying capacity, flight range of miles, wingspan of 10cm and lifetime of 2-3 weeks, Manduca sexta makes a wide range of applications for these devices possible.

Figure 1: The life-span of Manduca sexta during the metamorphic development and the results of insertions done at various stages of metamorphosis.

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE JAN 2008 2. REPORT TYPE

3. DATES COVERED 00-00-2008 to 00-00-2008

4. TITLE AND SUBTITLE MEMS Based Bioelectronic Neuromuscular Interfaces for Insect CyborgFlight Control

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Cornell University,School of Electrical and Computer Engineering ,Ithaca,NY,14853

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited

13. SUPPLEMENTARY NOTES 21st IEEE International Conference on Micro Electro Mechanical Systems (MEMS 2008), Tucson, AZ,January 2008, pp. 160-163

14. ABSTRACT This paper reports the first direct control of insect flight by manipulating the wing motion via microprobesand electronics introduced through the Early Metamorphosis Insertion Technology (EMIT). EMIT is anovel hybrid biology pathway for autonomous centimeter-scale robots that forms intimate electronic-tissueinterfaces by placing electronics in the pupal stage of insect metamorphosis. Our new technology mayenable insect cyborgs by realizing a reliable control interface between inserted microsystems and insectphysiology. The design rules on the flexibility of the inserted microsystem and the investigation towardstissue-microprobe biological and electrical compatibility are also presented.

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT Same as

Report (SAR)

18. NUMBEROF PAGES

4

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

2. EXPERIMENTAL PROCEDURE AND RESULTS Microprobe Design

In the case of flight muscle actuation, the main flight powering muscles are located in the dorsal-thorax of the Manduca sexta (Figure 2) where electronic implants can be located. The dorsovental and dorsolongitudinal muscle groups move the wings by changing the conformation of the thorax, which supplies the mechanical power for up- and downstrokes. The alternating relaxation and contraction of these muscles create the alternating up- and down-strokes hence the flight. Therefore, the designed probe should target actuating these muscle groups.

Probe flexibility is a required design property for wider probe geometries, since the probe has the potential to affect the biomechanics of muscle contraction. Conversely, muscle contraction can cause impact and high strain damage to the probe. We had previously demonstrated a microsystem with all-silicon probes [7]. Here, polyimide was selected as the base material of the probe due to its flexibility and biocompatibility. Moreover, fragility and delicacy of the probes while handling during the insertion and throughout the experiments is less of a concern with a flexible polyimide probe. The aimed experimental protocols consist of tethered setups where insect flight muscle is actuated through the flexible wires, as well as non-tethered setups where there are no attached wires and free-flight of insect can be realized. We designed and manufactured a flexible probe that can work with both setups (Figure 3B). The microsystem for autonomous control of the probe electronics can be seen in the same figure and consists of three parts: power, probe and control layers. The power layer (Figure 3D) is comprised of two coin batteries and a slide-switch positioned on a printed circuit board (PCB). Each battery has an energy capacity of 8mAh and weighs 120mg. Conductive adhesive was used to attach the batteries to the platform. The control layer (Figure 3A) is-

Figure 2: Cross-section (A) and illustrated diagram (B) of the flight muscles powering the up- and down-stroke of Manduca sexta wings.

Figure 3: The microsystem including microprocessor (A), flexible probe (B), silicon probe (C) and battery unit for power (D), the close-up view of the tip in (E) with the hole for muscle growth, the flexibility of the probe (F) and the assembled system (G). an 8×8mm2 PCB holding the microcontroller (Atmel Tiny13V) and an LED. The microcontroller was electrically connected to the PCB via flip-chip bonding. Wire-bonding was used to connect the PCB to the probe layer. The microfabricated silicon probe is sandwiched between these two layers (Figure 3G). The overall system has dimensions of 8×7mm2

and total mass of 500 milligrams.

The flexible probe can also be used in tethered setups by utilizing a FFC/FPC connector (Figure 10). All-silicon rigid probes, which provide higher stiffness for narrower cross-section enabling higher density probing, were also fabricated and tested (Figure 3C). Microprobe fabrication

Flexible PCB technology was used to deposit 18µm of Cu layer on 100µm thick Kapton-polyimide base material (Figure 5). Cu traces were coated with 20µm of LPI soldermask for insulation, except for the locations of the excitation/recording pads. 3µm of Electroless Nickel and Immersion Gold (ENIG) layer was deposited on the pads for biocompability. Each probe has a width of 400µm and each actuation pad is 75×75µm2 (Figure 4).

Figure 4: SEM image of the flexible-probe tip with expanded image of the ground and actuation pads

Figure 5: Cross-section (A) and description (B) of the layers used in the fabrication of the flexible probe

The relative stress between the implant and the tissue

was minimized by matched flexural rigidity (37.5N/m). For an all-silicon probe with a similar stiffness to the flexible probe, a silicon thickness of 30 µm would be required. Insertion results

The probe based microsystem platforms were inserted to the pupae 7 days before emergence (Figure 6). At this time, a thin thoracic skin is formed under the cuticle of the pupae. If inserted in earlier stages, the fluidity of the tissue prevented adequate sealing around the insert. When inserted later, some of the preformed muscle was damaged, leading to an inefficient bioelectronic interface. In addition, the flexible probes buckle and cannot be positioned to the targeted muscle groups when the thoracic skin is thicker.

Adaptation of probes by the muscle was highly maximized as the muscle grew around and through the hole of the probes (Figure 7iii), as observed under the microscope. Cuticle healing, therefore sealing, at the insertion points can be seen in the same figure (Figure 7i), both of which are indications of structural integration during metamorphosis.

When the probes were extracted, considerable tissue

was also removed in pupae-inserted probes, in contrast to adult inserted probes. Moths with inserted probes emerged with a success rate of 90% and have been electrically actuated.

Testing the tissue-probe coupling

The electrical coupling between the probe and the tissue was inspected before actuating the wing muscles using two

Figure 6: Pupal stage insertion (i) and successful emergence (ii). The microsystem platform on (ii) is held with tweezers to show wing opening of the moth. The X-Ray image of the dotted part (A) shows the probes inserted into the dorsoventral and dorsolongitudinal flight muscles. CT images (B) show components of high absorbance indicating tissue growth around the probe.

Figure 7: The crossection of thorax near the probe with explanatory schematic (ii) of thoracic flight muscles. Cuticle sealing (i) and muscle growth (iii) around the probe indicates integration by the body. (dl: dorsolongitudinal flight muscle, dv: dorsoventral flight muscle, see Figure 2) methods: (a) recording the muscle potentials during wing flapping, and (b) measuring I-V curves across the different probes in a tethered flight set-up (Figure 10). Probes that failed in any of these tests were discarded before the wing actuation experiments. The muscle potential recorded from the inserted probe was regarded as an indication of the goodness of the probe operation (Figure 8). The observed inter-spike duration is consistent with the wing flapping rate of moth (20-25Hz).

Rarely, the metal pads of the probe failed due to a currently unknown failure mechanism. Typical I-V curves of good coupling and such a failed probe can be seen in Figure 9. The actuated muscle fibers between the probe pads can be modeled with a simple 3 element RC network (Figure 9). Here RF denotes the resistance of intra- and extracellular fluids whereas RL is the leakage resistance of the membrane and CM is the membrane capacitance. The measured I-V curves (at DC) give the approximate addition of RL and RF. The resistivity values obtained from this sum (see Figure 9) for the good probes are in good agreement -

Figure 8: Muscle potential recorded from the dl muscles (see Figure 2 for dl) during wing-flapping. Observed spikes disappeared immediately with the recess of wing flapping.

Figure 9: The I-V curves of each electrical pad (measured with Keithley-4200) and the RC network modeling the muscle between the pads. The fitted lines and calculated resistivities are given in the table. Channel 4 (shown separately in 4th quadrant) has poor bio-electrical coupling. with the range reported for skeletal muscle in the literature (300-500 Ω⋅cm) [8-10]. The failed probes, however, reads abnormally reduced resisitivity values. Actuation of Flight Muscles

Phased actuation of probes with biphasic pulses allowed us to control wing motion highly selectively. Upstroke and downstroke actuation on “one” or both wings were demonstrated with power consumption of as low as 10 microWatts. By tethering the moth, we were able to affect the direction of insect flight by controlling the motion of the wing. Figure 10 shows unilateral up- and down-stroke evoked actuation of the wings. The wing actuation and direction of flight can be best seen in movie format [11].

Figure 10: The evoked up- and downstroke of a “single” wing obtained by applying 5V pulses to the indirect flight muscles (snapshots from the recorded movie). Under natural conditions, moths flap both wings together.

3. CONCLUSION We demonstrated a reliable hybrid tissue-electronics

interface in insects that provides flexibility against tissue movement. Inserting the probes at an early pupal stage ensures that the tissue grows around the probes for a highly natural implant. We also showed down- and up-stroke actuation of each wing separately, through which we were able to affect the flight direction of Manduca sexta. This work paves the way for future engineering approaches to utilize the bioelectronic interfaces especially for realizing insect cyborgs. 4. ACKNOWLEDGMENT

The authors would like to thank the members of the SonicMEMS Laboratory, especially Ayesa Paul and Abhishek Ramkumar for useful discussions and help during the experiments, the Beckman Institute of UIUC for being able to use their CT imaging facility and Janice Beal of Boyce Thompson Institute for rearing and supplying the moths. This work was fully supported by DARPA HI-MEMS program. The facilities used for this research include the SonicMEMS Laboratory, The NanoScale Science & Technology Facility (CNF) and The Nanobiotechnology Center (NBTC) at Cornell University. 5. REFERENCES

[1] G. Taubes, “Biologists and Engineers Create a New Generation of Robots That Imitate Life”, Science Magazine, April 7 2000, pp. 80-83, 2000. [2] C.P. Ellington, “The novel aerodynamics of insect flight: applications to micro-air vehicles”, The Journal of Experimental Biology, 202, pp.3439-348, 1999. [3] P. Mohseni, K. Nagarajan, B. Ziaie, K. Najafi, and S. B. Crary, “An ultralight biotelemetry backpack for recording EMG signals in moths,” IEEE Trans. Biomed. Eng., vol. 48, no. 6, pp. 734-737, June 2001. [4] J.R. Riley, “The flight paths of honeybees recruited by the waggle dance”, Nature, 435 (12 May 2005), pp. 205-207, 2005. [5] J. Mavoori, B. Millard, J. Longnion, T. Daniel, C. Diorio "A Miniature Implantable Computer for Functional Electrical Stimulation and Recording of Neuromuscular Activity", IEEE BioCAS 2004, Singapore, October 2004. [6] A. Paul, A. Bozkurt, J. Ewer, B. Blossey, and A. Lal, "Surgically Implanted Micro-Platforms in Manduca-Sexta", Solid State Sensor and Actuator Workshop, Hilton Head Island, pp. 209-211, 2006. [7] A. Bozkurt, A. Paul, S. Pulla, A. Ramkumar, B. Blossey, J. Ewer, R. Gilmour, A. Lal, "Microprobe Microsystem Platform Inserted During Early Metamorphosis to Actuate Insect Flight Muscle," IEEE Conference on Micro Electro Mechanical Systems (MEMS 2007), Kobe, JAPAN, pp. 405-408, 2007. [8] E. Zheng, S. Shao, J.G.Webster, “Impedance of skeletal muscle 1 Hz to 1 MHz”, IEEE Trans. Bio. Eng., vol. 31, pp.477-481,1984. [9] N. Sperelakis, T. Hoshiko, “Electrical Impedance of Cardiac Muscle”, Circ. Res., 9, pp.1280-1283, 1961. [10] N. Sperelakis, C. Sfyris, “Impedance analysis applicable to cardiac muscle and smooth muscle bundles,” IEEE Trans. Biomed. Eng., vol. 38, pp.1010-1022, 1991. [11]http://sonicmems.ece.cornell.edu/publications/movies/MEMS08.wmv

Micro Air Vehicles - Special Challenges

Documents