Transcript Detailed Design Review

Detailed Design Review
P15043
1
P15043
Detailed Design Review
12/9/14
Agenda








Concept Review and Updates
Electrical Detailed Design
Mechanical Detailed Design
Ergonomic Capability Analysis
Manufacturing Cost Analysis
Updated Bill of Materials (BOM)
MSD II Schedule
MSD II Roles and Responsibilities
P15043
Detailed Design Review
12/9/14
2
Review: Proposed Concept
Actuated Buttons
What we like about it:
• Easy to interpret feedback
• Direction is clear (rollers are not intuitive)
• Not significant internal design deviation (motors, wiring)
• Feasible in two semester schedule
• Does not require expensive technology (i.e. Bluetooth)
Trade-offs
• More moving parts
• Less flexibility with user grip
P15043
Detailed Design Review
12/9/14
3
Finalized Draft Model
P15043
Detailed Design Review
12/9/14
4
Cane Prototype Details
•
•
•
•
•
•
Aluminum cane that folds into equal sections
Total Length = 54.5”
Handle Length = 11”
Rod Length = 45.5”
2” overlap between cane and handle
Folding Mechanism: Bungee Cord
P15043
Detailed Design Review
12/9/14
5
EE Detailed Design Agenda
•
•
•
•
•
•
Microcontroller Functionality and Design
Accelerometer Implementation
Sensor Selection and Design
Electrical System Power Management
Motor Control Design
Test Plans
P15043
Detailed Design Review
12/9/14
6
Microcontroller Functionality
P15043
Detailed Design Review
12/9/14
7
High Level System Architecture
*Only one converter will be connected to
the battery at a given time.
P15043
Detailed Design Review
12/9/14
8
Microcontroller Schematic
• Atmega328
• Accelerometer in
microcontroller
schematic
P15043
Detailed Design Review
12/9/14
9
Accelerometer
•
Tracking Angular
Displacement
•
Tracking Angular
Acceleration
P15043
Detailed Design Review
12/9/14
10
Accelerometer Cont.
KXD94 series
•
Small, multiplexed analog output, High shock survivability, Low
Power Consumption
+- 10g’s at 200 mV/g
•
•
To track the position of the cane
Eliminates the need of two sensors
•
•
•
P15043
Reduces manufacturing cost, reduces total added weight and
increases space on the cane
Detailed Design Review
12/9/14
11
Sonar Sensor
LV-MaxSonar®-EZ2™: High Performance
Sonar Range Finder
• 5 V input capable
• Detects 3” objects up to 9.25 ft away
• Reads distance every 50 mS (20 Hz)
• Low-cost solution
P15043
Detailed Design Review
12/9/14
12
Decision to reduce to one sensor
 Individual Sensor Detection Range Geometry
P15043
Detailed Design Review
12/9/14
13
Decision to reduce to one sensor
 Maximum Detection Range with Two Sensors
• Mounting two sensors
onto the cane without
overlapping the sensor
beams creates a total
detection range of
178°
*Assuming the user sweeps the cane
90°
P15043
Detailed Design Review
12/9/14
14
Decision to reduce to one sensor
 Variable Detection Range with Two Sensors
• Overlapping range
beams allows for total
detection range to
vary between 134° and
178°
• More overlap means
smaller range
P15043
Detailed Design Review
12/9/14
15
Decision to reduce to one sensor
 Fixed Detection Range with One Sensor
• Limits total range
to 134°
• Engineering
Requirement: > 90°
• Decreases materials
cost
• Decreases current
draw from battery
P15043
Detailed Design Review
12/9/14
16
Battery
Inheriting P14043’s TENERGY Lithium Ion batteries, Li-ion 18650
3.7V 2600mAh
Reducing the number of batteries from two to one; using a boost
converter to meet voltage requirements.
•
•
This solution should have the system running for around 10 hours.
•
•
•
P15043
2600 mAh/300 mA ~ 9 h
200 mA is the maximum predicted current draw, also at full operating power
Detailed Design Review
12/9/14
17
Power Management: Battery Charging
•
Buck Converter LMR10515: Charging IC
P15043
Detailed Design Review
12/9/14
18
Power Management: System
•
Boost Converter from battery to system
P15043
Detailed Design Review
12/9/14
19
Microcontroller H-Bridge Design
•
•
Two DRV8839DSSR
Integrated Circuits
Allows for both forward
and reverse motor
control
P15043
Detailed Design Review
12/9/14
20
Test Plans
Battery Test Plan
•
Heat, power disipation, operation time over an 8-hour interval
•
Sensor Test Plan
•
Sensor accuracy at different ranges/degrees
•
Accelerometer Test Plan
•
•
Accuracy and position at different accelerations
• Buck and Boost Converter Test Plan
• Provide necsessary power output to components
P15043
Detailed Design Review
12/9/14
21
Tentative PCB Layout
P15043
Detailed Design Review
12/9/14
22
Moving Forward – MSD II
•
•
•
•
Focus on individual component functionality
Complete Test plans
Implement algorithm using development board to verify
feasibility
Build Prototype
P15043
Detailed Design Review
12/9/14
23
Questions?
P15043
Detailed Design Review
12/9/14
24
Cane Drawings
P15043
Detailed Design Review
12/9/14
25
Inside Placement
Legend
>Green – Micro Processor
>Red – Battery
>Blue – Rollers
>Yellow – Motors
>Orange – Handle Cap/Switch
Length: 11.00 inches
Outside Diameter: 1.30 inches
Inside Diameter: 1.10 inches
P15043
Detailed Design Review
12/9/14
26
Snapshots of Certain Pieces
Sensor Cover
P15043
Detailed Design Review
Sensor Wire Housing
12/9/14
27
Detachable Center
>Easier access to battery
>Necessary for prototype
manufacturing
>Easy access to repair inside
28
Benchmark Material
 Why we chose ABS?
 Great material properties for the application being used
 Easily available material
 Does not expand as much as Bridge Nylon
Filament
PLA
Polyactic Acid
ABS
Acrylonitrile butadiene styrene
Bridge Nylon Nylon
P15043
Density (lb/in^3)
0.043714
0.0376
0.041546
Weight per Pack (lb) in^3 per Pack
2
45.75193302
2.2
58.5106383
1
24.06970587
Before Drawing Estimation
Maximum Handle
Volume w/ extra
20.90431281
in^3
After Drawing Estimation
Creo Drawing
Handle Volume
11.1094
in^3
Detailed Design Review
Price per Pack
$
48.00
$
48.00
$
24.99
Distribution Company
Maker Bot
Maker Bot
taulman 3D
**Can print up to 5 handles with one
spool of filament
12/9/14
29
Motor Design Change
Original Design: Feedback oscillated in and out of cane handle via
linear actuator
Problem: A solenoid was the only motor type that was the right
size and within our budget. Solenoids have multiple drawbacks
including constant current draw and excessive heat build up.
Resolution: Decided to use P14043’s gear motor and have feedback
oscillation be from side to side.
Video of side to side
feedback oscillation
P15043
Detailed Design Review
IMG_1598.MOV
12/9/14
30
Roller Placement and Cane Mock-Up
 A feasibility analysis was done on the mechanoreceptors in the
hand to determine what areas would maximize tactile feeling.
 Different hand sizes were also taken into consideration to
maximize the range of potential users.
 Buttons were placed on a cardboard mock-up
to simulate movement and feedback placement.
P15043
Detailed Design Review
12/9/14
31
Roller Placement and Cane Mock-Up
 Multiple group members tested the button placement to see if
they could comfortably reach the buttons.
 The button placement was good, but the designed was changed
from buttons to rollers. This will decrease abrasion to the user’s
hand.
 A roller with a nylon body was chosen.
P15043
Detailed Design Review
12/9/14
32
Roller Feedback
Motor Arm
 Torque Calculation: 𝑡 = 𝑟𝑥𝑓 = 𝐿𝑎𝑟𝑚 ∗ (𝑃𝑔𝑟𝑖𝑝 ∗ 𝑆𝐴)
 Estimated Torque on Arm: 16.7 oz-in
 Max Torque Motor Output: 70 oz-in
 Motor Speed = 40 rpm (based on P14043)
 At 40 rpm, Motor Torque ≈ 42 oz-in
 Angle of Oscillation: ≈30-35° (Mock-up)
 Period, T=0.25 (P14043)
*Calculations shown on RIT Edge
P15043
Detailed Design Review
12/9/14
33
Haptic Feedback Test Plan
Engr.
Spec. #
S2
CDP4
Specification (description)
Signal detection of obstacles via haptic feedback
(horizontal and vertical motion in handle)
Grip stress handle can withstand and still have
feedback work
Engr. Spec.
#
S2
CDP4
Unit of Measure
Ideal
Value
Marginal
Value
Binary
-----
----
Pa
5 Psi
3 Psi
Instrumentation/Equipment
Test subjects and an obstacle course
Test subjects, weights, and grip force measuring device (in ISE lab)
Will consist of 4 Phases: Preliminary Motor Test, Feedback Response Test, Obstacle
Response Test, and ABVI User Test.
P15043
Detailed Design Review
12/9/14
34
Phase 1: Preliminary Motor Test
 Run the motor at 40 rpm with desired oscillation pattern (30
degrees, T=0.25s) with roller arm attached.
 Apply 5 psi pressure on the roller to verify the motor still works
under this condition. (To aid with this use the grip force
measuring device in ISE lab for pressure reference).
 Verify motor arm stays securely attached to motor during entire
test.
 If the motor and arm assembly pass this test move on to Phase 2.
P15043
Detailed Design Review
12/9/14
35
Phase 2: Feedback Response Test
 Install motor in the cane handle.
 Gather at least 5 volunteers. The volunteers should be blindfolded




and in a seated position during test.
Trigger right and left hand sides and record response time. Do at
least 10 trails on each person and record data.
Interview volunteers and record their feedback.
Identify improvement opportunities. Repeat testing if changes are
made.
Once feedback is satisfactory, move to Phase 3.
P15043
Detailed Design Review
12/9/14
36
Phase 3: Obstacle Response Test
 Create an obstacle course and gather at least 10 volunteers. Blindfold




each person and instruct them to sweep cane through course.
For each person record the number of obstacles they avoided and
compare this to the number of total obstacles.
Get user feedback on how well it works.
Evaluate if haptic feedback is satisfactory or if adjustments are
necessary.
If feedback works well move on to Phase 4. If feedback needs
adjustments, make the changes and repeat Phase 3 Test.
P15043
Detailed Design Review
12/9/14
37
Phase 4: ABVI User Test
 Bring the cane prototype to ABVI to test on blind users in an




obstacle course.
Get qualitative feedback from blind users on what they like and
don’t like about it.
Use their suggestions to make any needed adjustments.
If major adjustments are needed, after the changes are
implemented, bring the cane back to ABVI to test again.
Have a team meeting to review all testing procedures and finalize
haptic feedback design.
P15043
Detailed Design Review
12/9/14
38
Questions?
P15043
Detailed Design Review
12/9/14
39
Ergonomic Capability Analysis
Will the additional weight cause too much strain in the user’s wrist?
 How large of a moment is an adult wrist capable of enduring?
 Male: X~N(3.78 Nm,1.03Nm)
 Female: X~N(2.43Nm,0.74Nm)
 Moment placed on the wrist is a function of user height
 Average Male: 1.76m
 Average Female: 1.62m
 Moment placed on the wrist when maximum allowable weight is added (1 lb):
 Male: 2.18Nm, 94% capable
 Female: 2Nm, 72% capable
 Moment place on the wrist when only 0.5 lbs are added:
 Male: 1.45Nm, 98.8% capable
 Female: 1.34Nm, 93% capable
P15043
Detailed Design Review
12/9/14
40
Manufacturing Costing
Requirement: Manufacturing Costs < $125
Manufacturing Cost=
Materials + Labor+ Overhead + Misc. (fixtures, G&A)
Manufacturing “Elements”:
• Electronics (PCB, sensors)
• Handle (Handle production, motor and electronics assembly)
• Entire cane
41
P15043
Detailed Design Review
12/9/14
ABVI Manufacturing Costing
 Standard Labor Rate=$10.50/hr
 Manufacturing Cost =
(Material Cost + Labor*) + 11%
• *Time required as determined by time studies
• 11% is the assumed general and administrative rate
• Excludes fixtures, overhead and waste factors
• Materials from an ABVI standpoint would be:
• Handle assembly
• Collapsible Cane
42
P15043
Detailed Design Review
12/9/14
Updated Budget
P15043
Detailed Design Review
12/9/14
43
Updated BOM
Part #
Part Name
1
EZ2-MB1020
Description
Sonar Sensors
Qty
1
Units [number
sold in one
pack]
1
Photo
Bulk Unit
Cost
(USD)
$ 22.36
$
27.95
$4.01
$15.95
$
11.96
$
31.90
$3.95
$
3.88
$
3.88
$
7.76
McMaster-Carr 5-7 Business $
10.57
$
10.57
13.65
$
13.65
$
6.45
$
6.79
Source /
Company
298:1 Micro Metal
Gearmotor HP
2
1
Bearing for haptic
2
1
1
25
Pololu item #: 994
3
Unit Cost
(USD)
sparkfun.com/ 2-3 Business $
MaxBotix
2
Lead Time
Shipping
Days
http://www.polol 3-5 Busness
u.com/product/9
Days
94
Amazon
27.95
Cost
1-2 Months
feedback (button)
INA K3X6X7TN
4
3mm Diam, 40 mm long
Days
91585A076
5
High Strength Al bar
1
1
McMaster-Carr 5-7 Business $
1/2ft x 1" x 1/2"
Days
89215K425
6
Accelerometer
1
1
1
1
digikey.com
$
6.45
McMaster-Carr 5-7 Business $
6.79
$
3.23
KXD94
7
Double Shielded Bearing
Days
7804K124
P15043
8
Tenergy Li-Ion 18650
Battery
1
1
Amazon
$
12.50
$
12.50
$
12.50
9
ABS Filament
ABS - Filament
1
1 (2.2 lb pack)
MakerBot
$
48.00
$
48.00
$
48.00
10
LMR10515X
Buck Converter IC
1
1
digikey.com
$
1.31
$
0.69
$
1.31
Detailed Design Review
Free?
12/9/14
44
11
TPS61252
Boost Converter IC
1
1
digikey.com
$
2.48
$
1.36
$
2.48
12
ATMEGA328
uController IC
1
1
digikey.com
$
3.51
$
2.08
$
3.51
13
DRV8839DSSR
H-Bridge IC
2
1
digikey.com
$
1.62
$
0.88
$
3.24
14
10k Resistor
3
$
-
15
60.4k Resistor
1
$
-
16
6.65k Resistor
1
$
-
17
768k Resistor
1
$
-
18
243k Resistor
1
$
-
19
1M Resistor
1
$
-
20
1k Resistor
5
$
-
P15043
Detailed Design Review
$3.51
$0.00
$0.00
12/9/14
45
21
4.7u Inductor
1
$
-
22
1.5u Inductor
1
$
-
23
4.7u Capacitor
1
$
-
24
10u Capacitor
2
$
-
25
22u Capacitor
1
$
-
26
100p Capacitor
1
$
-
27
0.1u Capacitor
7
$
-
28
LED
1
$
-
29
MicroUSB port
1
$
-
30
Collapsible Aluminum
1
$
16.49
Ambutech
1 wk
$
16.49
$
13.09
$0.00
$0.00
$0.00
$0.00
Cane
P15043
Detailed Design Review
12/9/14
46
MSD 2
Schedule
P15043
Detailed Design Review
12/9/14 47
MSD II Roles and Responsibilities
Member
Major
Contact
Role
Emeka Akpaka
Electrical Engineering
[email protected]
Team Relations Lead
Kayla Cole
Industrial Engineering
[email protected]
MSD I Project Lead
Lindsay Johnson
Industrial Engineering
[email protected]
MSD II Project Lead
Justin LaMar
Electrical Engineering
[email protected]
Edge Coordinator
Christine Lochner
Mechanical Engineering
[email protected]
MSD I Tech Lead
Nick Stewart
Mechanical Engineering
[email protected]
MSD II Tech Lead
P15043
Detailed Design Review
12/9/14
48
Questions?
P15043
Detailed Design Review
12/9/14
49
APPENDIX
P15043
Detailed Design Review
12/9/14
50
Engineering Requirements
Importance
Source
9
CR1
9
CR1
3
CR2
3
CR3
3
CR4,
CR5
3
CR6
3
CR7
3
CR9
9
CR10
3
CR12
3
CR12
9
CR12
51
Unit of
Measure
Ideal
Value
Comments/Status
Degrees
90
Will be achieved by a combination of the
user's sweeping motion and 2, 25 degree
range sensors
Binary
Pass
Lbs.
1
Decrease amount of visible hardware by 50% compared to P14043
Pieces
10
8 hour rechargeable battery (minimum battery life)
Hours
8
Collapsible into 8-10" sections
Inches
8
USD
125
Minutes
1
Feet
10
psi
5
Binary
Pass
in
1.3
Function
System
Operation
System
Operation
System
Portability
System
Assembly
System
Operation
System
Portability
System Cost
System
Usability
System
Operation
System
Operation
System
Structure
System
Structure
P15043
Engr. Requirement (metric)
Provide 90 degree detection range in front of user
Signal detection of obstacles via haptic feedback (motion in handle)
Adds no more than 1 lb. to standard white cane
Manufacturing cost $125 or less
Keep cane collapse/re-open time less than 1 minute
Horizontal detection range
Maximum pressure
Handle contents fit within handle mock up envelope
Maximum handle grip diameter
Detailed Design Review
Less small parts would improve the
manufacturability of the product
Didn’t want to stall motor
Research on typical cane diameters
12/9/14
Critical Design Parameters
Number
Parameter
Unit of Measure
Ideal Value
CDP1
Cane handle outer diameter
in
CDP2
CDP3
CDP4
CDP5
CDP6
CDP7
CDP8
CDP9
CDP10
CDP11
CDP12
CDP13
CDP14
CDP19
CDP20
CDP21
CDP22
CDP23
CDP24
CDP25
CDP26
CDP27
CDP28
CDP29
CDP30
Handle length
Lag time between detection of an obstacle and feedback to the user
Handle grip stress
battery size
Total cane weight
Small number of pieces in handle assembly
Hollow space volume within handle
Input voltage of linear actuator
Voltage input type of motor
OD of rollers
Number of buttons
Dimensions of motor
Stall current of motor
Power for micro controller
Voltage input for the sensor
Sensor location in handle
Total horizontal detection range
Angle to mount the sensors
Total power draw
Size of battery
Weight of battery
Output of battery
Handle material
Lateral detection range
Wall thickness of handle
in
s
psi
mA-hr
lb
Number
in³
V
Binary
mm
Number
mm
A
V
V
Binary
Meters
Degrees
W
in
N
V
Binary
degrees
in
52
P15043
Detailed Design Review
Marginal Value
Owner
Involved
1.3
1.5
ME
ME
11
0.343
5
8 times the total current draw
11.5
<10
8.08
6
DC
6
2
10 x 12 x 35.27
1.6
5
5
<15
7.6
--------7
-------------
ME
EE
ME
EE
All
ISE
ME
ME
ME
ME
All
ME
ME
EE
EE
All
EE
EE
EE
EE
EE
EE
ME
EE
ME
ME
EE
ME
EE
All
ISE/ME
ME/EE
ME/EE
ME
ME/ISE
All
ME
ME/EE
3
10 ft from user
78.45
>= 5
Bridge Nylon
134 < x < 178
0.1
ABS
0.08
All
All
12/9/14
Subsystem Risk Breakdown
53
Subsystem
Risk
Motor
Medium
Buttons
Medium-high
Microcontroller
High
Battery
Medium
Sensor
Low
P15043
Detailed Design Review
12/9/14
Handle Repair Feasibility
 Goal: Make handle repair simple and easy for user or technician

Analysis: Main parts of system: Sensor, Battery, Micro-controller and
Motor
Instead of replacing the entire cane, small sections of the cane can
be replaced instead
 Solution: Design parts of the cane to be detachable or easily accessible. One
of the main parts that could break due to weathering is the sensor. The
current design was made with that in mind making sure that it could be
easily accessed for repair or replacement
**Examples of these design considerations will be shown in later slides**
54
P15043
Detailed Design Review
12/9/14
Linear Motor Research

A solenoid motor turned out to be the only type of linear motor we could find within our
price range and close to our dimension specifications.
Pros:
• Cost Effective
• Enables us to use original button design
Concerns:
• Constant current draw (will drain battery)
• Excessive heat build-up
• Solenoid housing is slightly larger than or preferences, so the handle design would
need to be slightly modified.
Conclusion:
Using a linear motor is not feasible, and the motor type and button motion must be redesigned.
55
P15043
Detailed Design Review
12/9/14
New Design Diagram
Bearing chosen
for rollers
56
P15043
Detailed Design Review
12/9/14
Button Placement Feasibility
Goal: Place buttons in a way to maximize tactile feedback feeling transmitted to the user
Background:
How the Blind Hold Their Cane
The blind hold their cane with their pointer finger extended down the flat side of the handle with the rest of their
fingers curled around it.
Mechanoreceptors
Mechanoreceptors specialize in sending tactile information to the brain.
Meissner’s Corpuscles:
 Directly beneath epidermis of fingers and palms
 Have rapidly adapting action potentials for shallow skin depression
 Suited for detecting low frequency vibrations and detecting textures moving
across skin
 Accounts for 40% of sensory nerves in human hand
57
P15043
Detailed Design Review
12/9/14
Button Placement Feasibility Continued
Figure 2: Two Point Discrimination Chart.
Image taken from source 3.
Figure 1: Distribution of Meissner’s
Corpuscles in the human hand. Image
taken from source 3.
P15043
58
Detailed Design Review
Figure 3: Indentation threshold for
different areas of the hand. Image taken
from source 3.
12/9/14
Button
Placement
Feasibility
Continued
Information Learned to Help Guide Design:

Meissner’s Corpuscles make up 40% of sensory nerves on hand.

Meissner’s Corpuscles have a high density in finger tips and an even distribution on
other areas
of the hand.

Meissner’s Corpuscles are good at detecting moving textured surfaces (consider
textured buttons)
59

Two point discrimination for the human palm is around10mm.

Two point discrimination for human fingers is around 5mm.

Fingers have a lower indentation threshold compared to the palm.
P15043
Detailed Design Review
12/9/14
Spatial Considerations
Handle Motors
Battery
Micro controller
60
P15043
Detailed Design Review
12/9/14
Sensor Selection Analysis

Based on rough measurements, the max sensor range length must be at least 8.75 ft.
 Thus, EZ3 and EZ4 were immediately ruled out.

Based on rough measurements of cane sweeping, the maximum angular displacement during a cane
sweep was determined to be about 45°.
 Thus, the sensor range angle can be no more than 45°.
The angles of the sensors’ ranges were determined, using half of the max range width and the
length at which said width is reached.
 Using one sensor on the cane, EZ0 provides a sensor range angle of 25°, which is desirable for the
lateral detection range. However, the height of the sensor range is about 7.5 ft., which provides a
range that is much too tall. Thus, EZ0 was ruled out.
 EZ2 provides a sensor range angle of 22°, and a sensor range height of about 4 ft. This is a desirable
height.
 Using two sensors, the EZ2 allows the sensor range angle to vary between φmin = 22°; φmax = 44°.
The optimal sensor range angle is assumed to lie within those values.

61
P15043
Detailed Design Review
12/9/14
Sensor Test Plan
 Read data from sensors using a development board
 Compare results using an oscilloscope
 Vary the detection object width/height and distance
 Testing of implemented algorithm using a cane (MSD II)
62
P15043
Detailed Design Review
12/9/14
High-Level Circuit Schematic
 High level design
 Emphasis on the input/output of the uController
 UART serial communication for Sensors
 GPIO pins for Motors and Accelerometer
63
P15043
Detailed Design Review
12/9/14
Decision to design a Printed Circuit
Board (PCB)
 Meeting with Carlos cemented the pre-production prototype idea
 Advantages:
 Cheaper in the long run
 Easy to modify/control all aspects of the uController functionality
 Disadvantages:
 Timely
 No previous experience
64
P15043
Detailed Design Review
12/9/14
MicroController Selection
 Key Selection Criteria:
 Low Cost
 Minimum of 8-bit core size
 5 V input voltage
 Analog-to-Digital Converters
Manufacturer
Unit Price(USD)
Core Processor
Core Size
Speed (MHz)
Number of I/O
Program Memory Type
Voltage Supply (V)
Data Converters
Microchip Technology
4.34
PIC
16-Bit
32
65
FLASH
2 - 3.6
A/D 16x10b
Texas Instruments
0.64
MSP430
16-Bit
16
10
FLASH
1.8 - 3.6
Slope A/D
Atmel
0.59
AVR
8-Bit
12
28
FLASH
1.8 - 5.5
A/D 8x10b
Atmel
0.86
AVR
8-Bit
20
16
FLASH
1.8-5.5
A/D 11x10b
Atmel
2.1
AVR
8-Bit
16
23
FLASH
4.5-5.5
A/D 8x10b
Selection: #5 – Atmel’s ATMEGA8
 Meets all of the selection criteria
 Chip commonly used for popular Arduino boards
 Troubleshooting resources readily available
65
P15043
Detailed Design Review
12/9/14
Power Management
 Powering:
 Two Motors, two Sensors, uController, and an Accelerometer
 Output of 5V and 150 - 200 mA for ~ 8 hours
 Choice of
 One battery with a boost converter
 Two batteries with a buck converter
66
P15043
Detailed Design Review
12/9/14
Buck/Boost Decision
 Buck Converter (decreases input voltage)
 Has large parts and takes up space on a PCB
 Having two batteries is not ideal (due to handle space)
 Boost Converter (increases input voltage)
 Also takes up room on a PCB
 Only needs one battery to operate
 Both provided the necessary power to components.
67
P15043
Detailed Design Review
12/9/14
Boost converter it is!
 Condensing the real estate in the cane to as little as possible is
invaluable
 Having one less battery also decreases the manufacturing cost
68
P15043
Detailed Design Review
12/9/14
Power Management Test Plan
 Plug 3.7 V power supply into converter
 Monitor the current draw from power supply (multimeter)
 Monitor the voltage output (oscilloscope)
 Monitor the current output (multimeter)
 Verify outputs match nominal values/simulations
69
P15043
Detailed Design Review
12/9/14
Accelerometer Selection
Kionix KXD94
•
•
•
•
70
P15043
Detailed Design Review
Small package
Low Noise
Low Power Consumption
Analog voltage output
12/9/14
Accelerometer Test Plan
 Read data from accelerometer using a developmental board
 Compare the data using an oscilloscope
 Vary the acceleration of the cane to determine the accuracy
71
P15043
Detailed Design Review
12/9/14
Proposed Design
 One battery with a Boost converter to power all of the
devices
 Two EZ2 sensors to acquire data from the environment
 Accelerometer to determine cane position relative to user’s
direction
 Using an AtmegA8 embedded system to implement our
algorithm and automate the entire system
72
P15043
Detailed Design Review
12/9/14
Collapsible Cane Research
73
P15043
Detailed Design Review
12/9/14
Risk Mitigation and Analysis
Importance
Effect
Severity
Likelihood
Risk Item
Cause
Actions to Minimize Risk
General
R1
Battery contact is compromised
Loss of power
Deflection of wire connection
2
3
6
Make sure all components that house wires are rigid and secure wires sufficiently for
cane movement
R2
User Muscle Fatigue
Pain/discomfort to user
•How hand grips on handle
•Weight distribution of cane
2
2
4
Ergonomics considered in design
Perform thermal analysis
R3
Over heating
Damage to system
Harm to user
Insufficient heat dissipation
1
3
3
R4
Cane malfunction
No feedback delivered to user
Component malfunction or damage
1
3
3
Design for redundancy
Multiple unconnected in the system
2
1
2
•Make system all one piece
•Create a way separate components can be stored together when not in use
Confusion and/or danger to user
•Sensor malfunction
•Broken connection
•Problem with program
1
3
3
Test prototype extensively
Sensors hit obstacles when cane is sweeping
•Damage to sensor
•Shift in sensor position
•Sensor falls off
Location of sensors on the cane
2
2
4
Attach sensors in the top region of the cane
Sensors get dusty/dirty
Malfunction
Environment encountered
1
2
2
State in user manual that sensors should be cleaned frequently
R5
Misplaced parts
User frustration
R6
Sensors relay incorrect information to feedback
R7
R8
Sensors
Handle
74
R9
Water damage
Ruined components
Not waterproof
2
3
6
•Minimize openings
•Put waterproof cover over feedback
R10
Loss haptic motion (when signal is send from sensors, feedback does not respond with
motion)
Feedback not given to user
•Disconnection of feedback mechanism and motor
•Burnout of motor
2
3
6
•Sufficiently secure roller to motor
•Do analysis to make sure torque is not too high for motor
R11
Haptic motion is unclear and not intuitive
•User confusion
•Learning curve to use cane
Haptic motion design
2
2
4
Do thorough testing to make sure haptic feedback relays information clearly to users
R12
Feedback is obstructed by clothing or jewelry (ex. Gloves)
Decreased feeling of feedback
Location where feedback comes in contact with the user
1
2
2
Brainstorm ways to minimize clothing/jewelry obstruction
R13
Motor vibrations harm user
Nerve damage
Magnitude of motor vibration (mm/s)
1
2
2
Do research on effects of vibration magnitude versus time of exposure. Ensure the
motor ordered is below the limit.
Degree of serviceability and ease part replacement
Defines if the handle can be fixed if a part breaks or if the user
needs to go out and buy a whole new cane
•Lack of access to inside components
•Can not remove/replace one part without removing/replacing
another
2
3
6
•Design handle with an easily removable insert that contains an organized array of all
handle components
•Use commercially available parts so that they can be ordered separately
R14
P15043
Detailed Design Review
12/9/14