manufacturing status review - colorado

Aaron Buysse

Chad Caplan

Matt Holmes

Colin Nugen

Kevin Rauhauser

Ryan Slabaugh

Rebecca Travers

Proximity Identification, chaRacterization, And Neutralization by tHinking before Acquisition (PIRANHA)

Manufacturing Status Review

8/24/2014 University of Colorado Aerospace Engineering Sciences 1

Team: Advisor:

Dr. Jelliffe Jackson

Customer:

Barbara Bicknell

Jeffrey Weber

•Overview

•Schedule

•Manufacturing

•Budget


Facts • 80k objects greater than 10cm

• ~60k in LEO

• Threat to assets and astronauts

• No active counter measure


Data of debris > 10cm in LEO, MEO, and GEO1

Purpose Schedule Manufacturing Budget


DETECT & RANGE THINK CAPTURE

Space Operations Simulation Center

6-DOF Robot

DCS

PIRANHA

POSITION FINDING

TOO BIG

TOO SMALL

FITS IN DCS

Debris Camera detects debris

Debris centered in camera FOV

Laser rangefinder ranges debris



Space Operations Simulation Center

Stage 1 Stage 2 Stage 3 PIRANHA: - Calculates, stores

debris relative position vector and DCS attitude

SOSC Robot: - Runs trajectory using pre-canned

attitude & ephemeris data stored on SOSC computer

SOSC Robot: - Runs trajectory using pre-canned attitude &

ephemeris data transmitted from PIRANHA

SOSC Robot: - Runs trajectory using pre-canned

ephemeris and real-time attitude data transmitted from PIRANHA

Ethernet Ethernet


MSR Week 8 Week 12

Mechanical Electrical Software Systems Test Class Milestone PIRANHA Milestone

• Converting Simulink/Matlab to C

• Hardware Component Tests

• PCB Power Board • “Hello World” on SBC • Machining Gimbal

Bracket

Validate Design in SOSC Mar. 27th

HWIL Phase I

Added – Consolidate software into 1 main program

Added SOSC Testing Feb. 11th

• Fit Check DCS to Robot • Run Scenario with DCS

attached • Transmit data structure from

SBC to SOSC

Software Phase II

HWIL Phase II

Verify Requirements


Choice of Voltage Regulators • Power Consumption Limits

• Proper Voltage

Trace widths • Board will draw a current of ≈5A in worst case

scenario • Equates to a trace width of 110mil on main line

Connectors to suit power needs • Input power was changed from header pins to

banana plugs because of expected current

Designed with SOSC Constraints in mind • 60V, 10A, 600W power supply limits

• Chose 15V input at a max of less than 5A



Power Board PCB • PCB designed, ordered, and arrived

• All other components ordered and arrived

Testing • 15V power supply, no loads on output pins

• Voltages on all output pins measured in reference to board ground pin (ground pin referenced to power supply ground)


Description Output Pin Expected (V) Measured (V)

Ground GND 0 0.00

Laser Rangefinder RNG 15 15.08

Single Board Computer SBC 5 4.90

Camera CAM 12 12.00

Servo 1 SER1 12 12.04

Servo 2 SER2 12 12.03

Fan for Heat sink FAN 12 11.95

FNC.5 - Testable in SOSC

CPE.4 - Integrate with SOSC

TSR.2.1, .3.1, .4.1 - Camera, Rangefinder, Single Board

Computer

Power Board Completed: Jan. 24th

Deadline: Jan. 27th

15VDC

Power Board • Individual components have been powered off

of board

• Full load test • All Components powered and running with

expected loads for the capture scenario

• Utilize main program as it is so far

• Determine Interfacing Connectors for distributing power

Electronics Box • Integrate Power board and Single Board

Computer

• Determine where holes need to be for mounting and cables


Power LED

Ground

15VDC

Camera


Integrate to electronics box: Mar. 11th


Camera

Raw Image

Image Processing

Servos

Debris Centered?

Rangefinder

Debris Edges

Calculate Rotation Angles

Size Characterization

Pan & Tilt Angle

Capturable?

ALERT DCS

Calculate Quaternions

Servo Interfacing

Data Packet

Camera Interfacing

Rangefinder Interfacing

Debris Range

Debris Size

no

yes

no

yes

FNC.2 – Detect debris 10-40cm in diameter at

20m

FNC.3 – Characterize debris by size to within

±2cm of actual size

FNC.4 – Calculate position of debris

relative to DCS

Software Flow Diagram of Capture Process: Driving Requirements

Software Trigger

Software Trigger Centroid

Detection


Image Processing: Detects edges of debris. • Tested & verified in C with SOSC test photos. • Obtained image processing filters from open source

computer vision software.

Centroid Detection: Determines debris centroid and number of pixels occupied by debris.

• Tested & verified in C with SOSC test photos. • Written by PIRANHA team members.

Size Characterization: Calculates size of debris given range.

• Tested & verified in C with SOSC test photos. • Written by PIRANHA team members.

SOSC Test Photo at 20m Zoomed debris after image processing

Debris

Size (cm) using C Size (cm) using MATLAB

Minimum Size 8.376 8.602

Maximum Size 11.04 9.218

Mean Size 10.03 8.902

Standard Deviation

0.792 0.229

Actual Size: 10.20 cm

Calculated centroid for all test photos in SOSC and calculated size to be within ±2cm of actual size.


Calculate Rotation Angles: Calculates rotation angles required to center debris in image.

• Tested & verified in C by comparing output to MATLAB results.

• Written by PIRANHA team members.

Calculate Quaternions: Calculates the quaternion of the DCS required to capture the debris.

• Tested & verified in C by comparing output to MATLAB results.

• Written by PIRANHA team members.

Convert all functionality software to C and perform Software Phase I testing by Feb. 3.

Calculate Rotation Angles

Centroid of debris throughout

capture process

Pan & Tilt Angles compared to

Simulink model

Calculate Quaternions

Range of debris Quaternions compared to

Simulink model Pan & Tilt Angles to range debris



• Integrated with FlyCapture C library • Used Linux native drivers

• Communicate over Ethernet with

packet size of 9000 bytes and packet delay of 6250 ticks (~5 × 10-5 sec.)

• Each image requires 2.8 Mb of memory

• Main code able to consistently trigger camera and acquire images

• Trigger laser rangefinder through RS-232 serial port

• Baud rate of 9600

• Outputs range in ASCII decimal format

• Main code able to consistently trigger rangefinder and acquire range measurements

• Command servos through RS-485 serial port

• Daisy-chained servos to command both simultaneously

• Commands inputs and outputs in hexadecimal format

• Able to command servos to angular position and get current position from servos in C

Two Dynamixel MX-64R Servos

Acuity AR1000 Laser Range Finder

Point Grey Flea3 GigE Camera

HWIL Phase I Testing - Integrate SBC with each hardware component individually by Feb. 18.


Task Deadline Status

Convert MATLAB code to C & SW Phase I testing 02/03/2014 Complete – 01/31/2014

HWIL Phase I Testing 02/18/2014 Complete – 01/28/2014

Integrate all functions into PIRANHA source code 02/18/2014 Partially Complete

SW Phase II Testing 02/25/2014 Incomplete

HWIL Phase II Testing 03/18/2014 Incomplete

Tasks moving forward • Integrate function to calculate quaternion of DCS required for capture into main code • Integrate decision making logic into main code • Software Phase II Testing – Verify that all functions operate as required when integrated into main

code • Integrate servo interfacing functions into main code • HWIL Phase II Testing – Verify software can interface with all hardware and is able to track debris



62 cm

90 cm

Heritage Debris Capture System

PIRANHA 20 cm

56 cm


Rangefinder

Camera

Software and Computing Hardware

2-DOF Gimbal Assembly

Servos

Gimbal System • 2-DOF

• Camera at center of rotation

Purchased Components • Rangefinder, camera, servos,

bearings, fasteners

Manufactured Components • 2 Gimbal Brackets

• Ti-6Al-4V 3D printing at Lockheed Martin


FNC.2: Detect debris at 20m

DES.4.2: Pointing & Tracking during entire capture scenario

CMP.2.1.1: Camera shall have 2DOF

independent of DCS

FNC.4: Pointing & Tracking Debris

CMP.3.1.1.3: Rangefinder rigidly attached to camera Tilt axis of

rotation

10.4 cm

7.1 cm

2.8 cm Gimbal Corner Bracket

4.8 mm thick

Camera/Rangefinder Bracket

14.6 cm

4 cm

4.4 cm

10.6 cm

Pan axis of rotation

Press fit bearings with pin

Laser beam

Tilt axis of rotation

Printed Bracket as Received

No Holes or Flange

Gimbal Brackets

• CAD models revised for printing

• Removed holes, camera flange

• Part requires post machining, assembly

• Printed brackets received from Lockheed Martin

• Defective bracket machined first – CNC mill

• Unfamiliarity with machining printed material

• Can assess dimensional accuracy of post-machined printed part

• SolidCam G-Code generated can be used for real bracket



Ball Bearing Press Fit into

Bracket

Camera Flange

Camera/Rangefinder Bracket Assembly Bracket After Machining

Holes Added

Front Edge Smoothed

Alignment Important Along Tilt

Rotation Axis

Tilt Rotation Axis

Gimbal System • Finish machining brackets, assemble system

Electronics Enclosure • Finalize mounting positions of components

• Will require addition of holes for fasteners, wires, fan

Mounting Plate • Aluminum plate material available

• Requires simple machining

PIRANHA/DCS Integration • Will need to add 4 holes to DCS side

• Mounting location determined to avoid bristles & structural components of DCS


Mounting Plate 3/16” Thick Al

Requires 12 Holes

4 Holes Added to Side of DCS

Electronics Enclosure Fan, Wire, Screw Holes

Added

Gimbal Corner Bracket

Bearing and Screw Holes

Mechanical Work Deadline: March 1st

Rangefinder

• Measured ranges within required accuracy of ±3mm

Camera

• Able to take image

• Still need to test with lens (ordered but not received)

Servos

• Tested unloaded and loaded

• Pointing error greater than ±0.14˚ required

• Possible solutions: apply correction to command, control law, higher voltage


Tapes 1m Apart

Rangefinder Aligned with Tape

Rangefinder Testing

Servo Testing

Bending Moment Case Lateral Loading Case

Servo Weight

Servo

Weight


Expenditures

Electronics

Hardware

Mechanical HW

Miscellaneous

Budget Left Over

Future Expenses

Poster Printing

Miscellaneous HW Additonal Funds

• After future expenses ≈ $900, Aero department provides $500 for travel

• We have ≈ $1,400 for travel to the AIAA student conference

Total Expenses: $3833.54

Estimate from CDR: $3397

Difference: $436.54

• There were a number of small purchases overlooked at CDR • We no longer can repurchase some large components if

damaged • Not a significant concern due to the level of understanding of

the hardware at this point


1 Bemner, Brian. "Cleaning Up the Final Frontier." Bloomberg Business Week. Bloomberg, 16 Jan. 2014. Web. 02 Feb. 2014.


Slide Intentionally Left Blank

Slides here after are backup.



Purpose: This test is to verify that the laser rangefinder’s output data can be collected and stored by the single board computer. This will help mitigate risk of PIRANHA failing due to the software algorithm being unable to access range measurements.

Diagram of Test Set-Up:

Summary of Measurements:

Required Materials/Facilities: Requires the PCM-9362 (SBC) with Linux and use of GCC complier, AR1000 (LRF), DC power supplies of 5V and 15V, Computer monitor,

Measurement How the Measurement is Verified Related Requirement

Laser Rangefinder output range stored on SBC

Compare physical range of LRF to data transfer range

CMP.4.1.1


-Using a C routine on the single board computer, pick off streaming range measurements at 6 Hz (fastest for rangefinder) -Some measurements unusable, and the rate decreased for dark objects LESSONS LEARNED - Original method of streaming range measurements is

difficult. - Investigate alternative to buffering a stream of

measurements

Unusable


• Purpose: Verify the rangefinder can accurately range targets within the tolerances specified at the ranges needed for testing in the SOSC. Test will mitigate possibility of failure due to defective rangefinder

• Diagram:

Distance = 1, 5, 10, 15, 20 meters

Acuity Accu Range 1000 - Rangefinder

Student Laptop



Required Materials/Facilities:

• Wall

• Meter tape

• Laptop capable of interfacing with rangefinder

• Steady surface for rangefinder

Measurement How the Measurement is Verified

Expected Accuracy Related Requirement

Range (m) Truth distance is measured with meter

tape

+/- 3 mm (AR1000 Data Sheet)

DES.2.1, DES.4.1


Measurement 1 (m) 2 (m) 3 (m) 4 (m) 5 (m) 20 (m)

Trial 1 0.951 1.951 2.953 3.953 4.950 20.102

2 0.951 1.951 2.953 3.953 4.950 20.102

3 0.951 1.951 2.953 3.953 4.950 20.102

4 0.951 1.951 2.953 3.953 4.950 20.102

Distance from

tape #1

0 1 2.002 3.002 3.999 - Tape #1

2 m

3 m

4 m

LESSONS LEARNED: -Rangefinder has precision stated in data sheet -Reasonable ranges returned at just over 20 m -Rangefinder model does not need to be updated. -Must calibrate range prior to operation with PIRANHA to get required accuracy.


-Need appx. 1.5° accuracy in servo pointing to capture with proper DCS attitude, and 0.24 Hz sample rate to complete turn at 3°/s

-2 Sources: Servo Pointing Error and turn timing error.

-Have theoretically 0.14° pointing accuracy and 1.59 Hz sample rate


Electronics Cost

Single Board Computer $404.01

Power Board $50.44

USB to RS-485 Converter $30

Electronics Enclosure $39.17

Power Board Components $35.18

Power adapter $4.95

Total $563.75

Purpose Solution CPEs Requirements Risks V&V Planning

Mechanical HW Cost

Bearings $21.66

Pins for bearings $12.57

Screws/Test eq. $20.86

Total $55.11

Miscellaneous Cost

FFR Printing $44.59

Total $44.59

Instruments Cost

Range Finder $1282.50

PointGrey Flea3 Camera $1085

Camera Lens $140

2 x MX-64 Servos $611.80

2 Servo Brackets $50.79

Total $3170.09

TOTAL = $3,833.54


Purpose: Verify that the algorithm used to calculate the appropriate attitude quaternion to

output to the SOSC robot does so correctly based on requirements. This test also will mitigate risk of the DCS capturing the debris in a manner outside of its capture alignment limits.

Diagram:

Rangefinder Model

DCS Attitude Quaternion Calculated

Euler Angle Servo Model

Logic

Capture Scenario Trajectory Model

Desired Attitude Quaternion

Compare

Quaternion Calculation Algorithm

Summary of Measurements: Required Materials/Facilities: -Computer which runs Simulink for preliminary test, or single board computer running

chosen environment for secondary test in C -Capture Model



Expected Result Related Requirement

Calculated DCS Attitude Quaternion

Capture Model Quaternion Turn begins within 1/6th second, final angle within

0.14 degrees

DES.4.1.1


Metric Experimental Value Expected Value

(Turn Begin Time) - (Ideal Turn Begin Time) 0.1668 [s] 1/6 [s]

(Final Angle) – (Ideal Final Angle) 2.9*10^-3 Less than 1*10^-2

Lessons Learned: -Function performs as expected based on the capture scenario model. -Set up for easy evaluation of errors. Timing of the scenario can be changed to simulate computation time, and range and measured camera angle errors can be introduced.

Q3


Purpose: Verify that the chosen servos (MX-64T) are capable of performing as required when subject to loading characteristic of the PIRANHA instrument.

Commanded Angular Position

Servo Rotation Rate Compare

Measurement How the Measurement is

Verified

Expected Result

Related Requirement

Servo Final Position

Comparison with servo command

Within 0.088

degrees

TSR.2.2


Diagram

Required Materials/Facilities: • Servo • Mass representing PIRANHA loading • Computer for communication with servo • Variable angle servo mount

Comparison of Commanded and Resulting Angular Position

Required Rotation Rate

(capture scenario model)

= Mass Characteristic of PIRANHA Load on Servo

= Rotation Axis

= Servo

= Vertical Reference

θ

Θ varied from 0 to 30 degrees


-2 cases, servo shaft bending moment loading and shaft lateral loading. -Lateral loading results most interesting -With loading incremented from 0 to 1.62 kg the servo was commanded to move to the position outlined in red.

Servo Shaft

Load [kg] Discrepancy Between Commanded and Measured Position [deg]

0 0.177

0.5 1.408

1.62 4.484

Commanded Position

LESSONS LEARNED -Must test the assembled system to understand servo response. -Possible solutions: Add bias to the control software, or increase servo voltage


𝑑𝑝 = 𝑟 tan(𝛼𝑚𝑎𝑥𝑡𝑠)

𝑑𝛽

𝑑𝑡= 𝛼 = Angular rate of change of 𝛽

𝛽 = Angle from DCS longitudinal axis to debris

𝛼𝑚𝑎𝑥 = maximum angular rate throughout the scenario = 0.06 °/𝑠

𝛾 = Half of total field of view of camera

𝑟 = Range to debris

𝑑 = Distance across an image at a specified range

𝑛𝑝 = Number of pixels across an image

𝑙𝑝 = Length of one pixel

𝑙𝑜 = Length across the debris

𝑡𝑠 = Camera shutter time

𝜌 = Smear ratio

𝑛𝑜 = Number of pixels across an object at a specified range

𝑑𝑝 = Distance a pixel moves through in

one picture

𝑑 = 2𝑟 tan 𝛾

𝑙𝑝 =𝑑

𝑛𝑝

𝜌 =𝑑𝑝

𝑙𝑝

𝑙𝑝𝜌 = 𝑟 tan(𝛼𝑚𝑎𝑥𝑡𝑠)

𝑡𝑠 =tan−1 𝑙𝑝𝜌

𝑟

𝛼𝑚𝑎𝑥

𝑛𝑜 =𝑙𝑜𝑛𝑝

𝑑 % 𝑠𝑚𝑒𝑎𝑟 =

𝜌

𝑛𝑜

𝑟

Camera View

Debris 𝑑

𝛾 DCS DCS

Debris

𝛽 𝑑𝛽

𝑑𝑡= 𝛼

Relative Attitude

Nomenclature

• 𝜌 = 0.2 (1/5 of a pixel is smeared before camera is able to take a picture) • Need 𝑡𝑠 = 5.4ms

• Percent smear of object pictured in figure

• Camera has a shutter time of 30𝜇𝑠 which is much faster

• In worst case, with 𝑡𝑠 = 5.4ms and a 10cm debris at 20m smear is only about 1.2% of the object.



Open Ethernet port

Connect to camera

Establish Receive Buffer

for Images

Create image structure in

image variable

Initialization procedure

Begin streaming

images

Send software trigger to

acquire image

Save image to receive buffer

Stop streaming images

Initialize camera context and

image variable

Raw Image

Trigger Camera

PIRANHA Main Code


Initialize pointers to buffer

Open RS232 port

Send trigger to acquire range

Read serial port

Convert range from ASCII

decimal to integer

• Triggering rangefinder to get range rather than picking range of stream of measurements to decrease processing time

Laser Rangefinder Interface Flow Diagram:

Servo Interface Flow Diagram:

PIRANHA Main Code

Range

Construct data packet

Send data packet to servo

Pan & Tilt Angles

Read serial port Convert angles to

hexadecimal Open

RS485 port

Angles

PIRANHA Main Code

• Sample data packet: {0xFF,0xFF, ID, Length, Instruction, Address, Low value, High value, Checksum}


Gimbal Corner Bracket


F

d

Beam Cross Section

Cantilever Beam

Max. Deflection (Ti-6Al-4V) = 0.55 μm Max. Deflection (6061 T6 Al) = 0.75 μm

Beam Cross Section

F

a b

l

Max. Deflection (Ti-6Al-4V) = 0.28 nm Max. Deflection (6061 T6 Al) = 0.41 μm

d

a b

l

Tilt Servo

Pan Servo


Tilt Servo

Pan Servo

d

d

T F

d

F

F

d

F T

Req. Torque (Ti-6Al-4V) = 1.77 Nm Req. Torque (6061 T6 Al) = 1.48 Nm

Req. Torque (Ti-6Al-4V) = 2.08 Nm Req. Torque (6061 T6 Al) = 1.83 Nm

At 11.1V, 3.9A, servo stall torque is 5.5 Nm (>2x any required torque)

Moment: 1.73 N-m

• Will order servos early for testing

• Will manufacture “panning” bracket if necessary • Bending moment will be held by

bracket

• Only torque on servo shaft


Panning Bracket

Bearings • Combination of ball

and thrust bearings • Support radial load

on servo shaft • Allow free panning

rotation

Pan axis of rotation


Purpose: Verify all voltage and current levels are correct to ensure correct operation of all

critical components to the project. Designed to mitigate risk of frying components before and while they are connected. Voltage levels will be checked prior to hooking up sensors to verify correct voltage regulation.

Setup:

Voltage and current measurements will be taken at the points marked in red. All voltage measurements will be in reference to ground.

15V ±2mV

Expected Value Camera SBC Servo Rangefinder

Voltage (± 2mv) 12V 5V 12V 15V

Max Current 0.21A 2.31A 1.00A 0.13A

Summary of Measurements :

Materials: • Multimeter • Oscilloscope (voltage regulation) • Power PCB • Electronic components in circuit


Purpose: Verify that camera gimbal system provides two degree of freedom motion with the pointing accuracy given by the component requirements. The test is designed to mitigate the risk of camera pointing errors by measuring the accuracy of the system performance.

Command

Position output from servo

Commanded Angular Position

Position Output from Servo

Compare


Verified

Expected Accuracy

Related Requirement

Vertical servo angular position

Comparison with input

± 0.088˚ CMP.2.1.1.1

Horizontal servo angular position

Comparison with input

± 0.088˚ CMP.2.1.1.1


Diagram

Servo Controller

Required Materials/Facilities: • Camera gimbal assembled according to design

• Two manufactured gimbal brackets • Two servos • Large mounting plate

• Servo controller (computer with USB to RS-485 converter)

• Communication cables (USB and RS-485)

Comparison of Commanded and Resulting Angular Position


Purpose: Verify PIRANHA is able to accurately calculate and store attitude and the relative position vector of the debris at a rate no less than 10Hz as it moves through a pre-determined STK scenario. This test will mitigate risks associated with PIRANHA not being able to determine the position of the debris relative to the DCS accurately, attitude of the DCS accurately, or at the required rate.



Expected Accuracy

Related Requirement

Debris relative position calculated by PIRANHA

Comparison with debris relative position & Attitude from STK scenario

± 0.045m DES 4.1 DES 4.2

Required Materials/Facilities: Diagram • SOSC Facility

• Ephemeris & Attitude data files • PIRANHA & DCS

PIRANHA

Debris Relative Position & Attitude Calculated

by PIRANHA

Debris Relative Position & Attitude from STK Scenario

Compare

Comparison of Debris Relative Position Vectors


Purpose: Verify PIRANHA is able to output an ephemeris & attitude data structure at a rate of no less than 10Hz. This test will mitigate risks associated with PIRANHA not being able to communicate with the SOSC robot.



Expected Accuracy

Related Requirement

Data transfer Rate

Analyze data post Ethernet transmission

Transfer rate > 10Hz

DES.5.1

Diagram

Required Materials/Facilities: • SOSC Facility • Ephemeris & Attitude data files • Ethernet cable • PIRANHA & DCS

Transmitted Ephemeris & Attitude File

Ethernet Received Ephemeris

& Attitude Compare

Comparison of Ephemeris & Attitude Data Structure

PIRANHA Ephemeris & Attitude Downlink


Purpose: To verify the capability of detecting an object in the S.O.S.C. environment. This authenticates that our image processing function can detect an object between 1 and 20 meters, with a diameter of 10 cm. This test verifies that the algorithm developed can satisfy FNC.2, DES.2.1, and TSR.2.1: Detection of 10 cm diameter at 20m with a camera.


Diameter

Range

Physical Test Set-Up:

Given diameter and range, should object be

detected?

Compare

Summary of Measurements: Required Materials/Facilities: -SOSC

-White sphere of 10 cm diameter -Computer running MATLAB, or eventually single board computer running C and chosen environment


Measurement How the Measurement is Verified Expected Result Related Requirement

Binary, Detected of Not? Physical dimensions of test set up Objects of 10 cm diameter are detected at a range of 20 m

FNC.2, DES.2.1, TSR.2.1


Purpose: To verify the software size characterization algorithm in the SOSC environment at

varying ranges. This was done utilizing a camera, rangefinder, and image processing software. Designed to mitigate risks of errors in size characterization during full scale test by adjusting variables in code to match SOSC environment.

Setup:

Summary of Measurements :

Measurement How it was completed

Debris Diameter String around circumference

Range to Debris Handheld Laser Rangefinder standing under camera and debris

Camera FOV Given by SOSC

Size of Debris Image processing algorithm

DES.2.1: Ability to detect 10 – 40cm diameter debris at 20m DES.3.1: Ability to determine size to ±2cm

Co

mp

are


Results: 10.2cm Diameter Debris at a range of 4.20m Calculated Size = 9.22cm

Test Distance [m] Calculated Size [cm]

1 20.30 9.0

2 4.20 9.22

3 20.12 & 19.82 8.63

4 15.16 8.87

5 10.00 8.93

6 5.46 – Materials: • MATLAB • Camera • Rangefinder

Facilities: • SOSC • Maneuverable

Robot Size characterization was within ±2cm requirement for each test Actual Size

of Debris Image

Processing Calculated

Size Compare

Processed Image


Purpose: Verify all software components, when put together, function in line with expectations

and requirements. The test will also provide additional insight into the possibility of PIRANHA failing due to software function interfaces.

Diagram: Image of Simulated Debris

(from camera and capture models)

Range (from range and capture models)

Camera Pointing Commands

Relative Vector to Debris

Quaternion for SOSC

Debris Size Characterization

Simulink Block containing PIRANHA “main”

Capture Scenario Model

Compare

Image of Simulated Debris

Compare

Summary of Measurements: Required Materials/Facilities: -Computer which runs Simulink for preliminary test, or single board computer running

chosen environment for secondary test in C -Camera, rangefinder, capture scenario models




Camera Pointing Commands Comparison with input Within 0.14 degrees of capture scenario geometry

CMP.2.1.1.1, CMP.4.3.1.1

Vector to Debris Comparison with input Angle within 0.14 degrees, magnitude within 0.045 m

CMP.4.3.1, DES.4.1

Quaternion to SOSC Capture Scenario Model, input Identical to capture scenario model

DES.4.1.1, DES.4.3

Debris Size Characterization Input Reflects simulated image of debris

DES.3.2


Purpose: This test is to verify that the camera can be trigger via software and an image can be stored on board the single board computer. This will help mitigate risk of PIRANHA failing due to the software algorithm being unable to access images.



Required Materials/Facilities: Requires the PCM-9362 (SBC) with Linux and use of GCC complier, Point Grey Flea, DC power supplies of 5V and 12V, Computer monitor


Image stored on SBC Image viewed on computer monitor CMP.4.1.1


Purpose: This test is to verify that a servo can be instructed via software to more to a position and then have the final position returned to the single board computer. This will help mitigate risk of PIRANHA failing due to the software algorithm being unable to point the rangefinder/camera, or read off angle measurements for the relative position vector.



Required Materials/Facilities: Requires the PCM-9362 (SBC) with Linux and use of GCC complier, MX-64 Servo, DC power supplies of 5V and 12V, Computer monitor


Send and receive data from servo Compare output position from servo to command position

Purpose: Verify the camera operates with the FOV and pixel count specified in data sheet. Test will mitigate the possibility of failure due to defective camera, or the incorrect FOV and pixel count entered into software algorithms

Diagram:


Image Processing Software for pixel

count

Known distance

Mea

sure

d d

ista

nce

to

ca

lcu

late

FO

V

Verify pixels and FOV


Required Materials/Facilities: • Wall

• PointGrey Flea 3.0

• Image Processing Software




Pixels Pixel count DES.2.1, DES.3.1

FOV (deg) Distance for field of view known at 20 meters

Depends on Lens DES.2.1, DES.3.1


Purpose: Verify that the control law, when connected with the camera, centers an object in the field of view in the manner stated by project requirements. The test will also help in mitigating the risk of PIRANHA failing due to inability of the debris to be ranged.

Object in FOV not Centered

Control Law Commands

Servos

Compare


Verified

Expected Accuracy

Related Requirement

Calculated Centroid offset from Field of View Center

Comparison with control law

software test

± 0.14˚ maximum

CMP.2.1.1.1 CMP.4.2.1.1


Diagram

Required Materials/Facilities: • Camera gimbal assembled according to design

• Servos, camera, rangefinder attached • Single Board Computer • Communication cables (USB and RS-485) • 2 white spheres, 10 and 40 cm in diameter • Black Sheet

Comparison of Object Position in FOV with Ideal

Servos Reposition

Camera

Object Moves in FOV

Center of FOV

FOV FOV


Purpose: Verify that in a capture-like scenario, PIRANHA is capable of tracking, ranging

characterizing debris, and outputs quaternion and ephemeris data in a form acceptable by the SOSC. The test will also verify that all hardware and software is interfacing correctly, and will help in mitigating risk of failure associated with running all software and hardware concurrently.

Diagram: Camera sees where debris is

on the table

Fishing Line to Move Debris

PIRANHA

Debris Size Characterization

Quaternion/Ephemeris Compared to required format

Table

Debris less than 1 meters verified to begin the DCS turn



Required Materials/Facilities: • Table with minimum dimensions 2x1 [m] • Paper on which to create position reference for debris, or graph paper • Secondary camera (808 # 16 camera already available, or cell phone) • Timer with visible digits • PIRANHA • Sphere to simulate debris, 10-40 cm • Fishing Line


Expected Accuracy Related Requirement

Relative Position Vector Known Relative Position via Camera and Table Markings

+/- 3 mm (AR1000 Data Sheet), pointing within 0.14 degrees

FNC.4, DES.4.1, CMP.4.2.1.1

Relative distance at which Quaternion Output Changes

Known Relative Distance via Camera and Table Markings

+/- 3mm DES.4.2

Ephemeris and Quaternion Format and Rate

SOSC Known Requirements Identical FNC.5, DES.5.1

Debris Characterization Simulated Debris Measurement

Reflects Test Case Ball Size CMP3.1.1.1, DES.3.2

manufacturing status review - colorado

Documents