S.N.A.K.E.: A Dynamically Reconfigurable Artificial Sensate Skin ...

More documents

Recommendations

Info

So in average, the strain gages consume a total of 30mA when turned on, which is why it was so important to gate them using the MOSFET. Total power consumption can be then as low or as high as: Low 300nA + 2.8µA + 2(700µA) + 4(320µA) + 4nA + 10µ A + 62µA + 175µA + 100nA = 2.927404mA High 2.4mA + 2.8µA + 2(700µA) + 4(320µA) + 4nA + 10µ A + 620µA + 175µA + 3mA + 6mA = 14.887904mA In actual tests, maximum power consumption of one node with the MCU running at full speed, all LEDs on at full power, both strain gages on and a laser aimed at the light sensor, was of 41mA. When everything was turned off however it consumed less than 2mA. If we now combine this with the possible number of nodes in a Skin Patch, we can see that a Skin Patch with 40 nodes at full power would consume 1.64A of current, but this case would be rare. On average each node consumed a total power of less than 9mA in regular operation, so a total of 40 nodes would consume roughly 360mA. Still, it is important to note that even though 9mA was the average current consumed, this is highly dependant on the application being run by the nodes since most of this will be dissipated by the LEDs. Their total current consumption can be easily brought down though, since they are driven by timer output pins, which means that they can be pulse- width modulated. Another thing to note is that power dissipated by components does not have any measurable effect in temperature sensor readings. This is mainly because the amplifiers and RGB LED, which are the most power-hungry elements, are placed so far from it that most of their heat dissipates to the surrounding air rather than to the substrate. Moreover, Kapton is not a good heat conductor so it would be extremely rare to see any effects in the temperature readings caused by heat dissipated from nearby components. 138
Finally, since power is readily available for Brains, and their power consumption is negligible compared to the total power that a large Skin Patch could in theory consume, this was not an issue and was therefore not calculated. 5.3.3 Cost The final metric used is the overall cost of the project. Cost was a big consideration in the overall design of the project because flex boards are expensive to fabricate and even more expensive to prototype, specially if multilayer boards and other fancy techniques are used. Without a doubt, the major cost component of the project was the fabrication of the node substrates which accounted for roughly two thirds of the total cost. 32 nodes were fabricated at a total cost of 3,500, so each node cost 110. Component prices for nodes and Brains can be observed in the Bill of Materials included in Appendix B. Two prices are listed in this table: the price used for the project and a bulk price for each component. This was done with the intention of providing a comparison between the cost of a prototype project like this one and its possible cost with large-scale manufacturing. It can be seen that the components for each node add-up for a total of 54 per node or about a third of the total cost. If these components are bought in large quantities though, the cost decreases to about half as much. Even substrates can be manufactured for about a tenth of the cost if they are fabricated in large numbers. Brains are much cheaper because they use standard fabrication techniques and relatively cheap components. Also the price can be seen in the Bill of Materials. 139
Page 1:
Author S.N.A.K.E.: A Dynamically Re
Page 5:
S.N.A.K.E.: A Dynamically Reconfigu
Page 9 and 10:
Table of Contents Abstract 3 Acknow
Page 11 and 12:
List of Figures 2-1 Created by MIT
Page 13:
E-1 OpenGL Control Header File . .
Page 17 and 18:
Chapter 1 Introduction “Imaginati
Page 19 and 20:
The current project, named S.N.A.K.
Page 21:
Hardware Each Skin Patch is compose
Page 24 and 25:
living tissue. Particularly in the
Page 26 and 27:
Figure 2-2: Leonardo’s Hand great
Page 28 and 29:
Artificial Sensate Skins elsewhere
Page 30 and 31:
2.1.2 Ultra-dense Sensor Arrays Thi
Page 32 and 33:
2.2 Flexible Circuits Although, not
Page 34 and 35:
2.2.2 Other Technologies Figure 2-6
Page 36 and 37:
human skin. The following design pr
Page 38 and 39:
(a) Skin Patches as data extraction
Page 40 and 41:
Figure 3-3: I 2 C Routing pattern T
Page 42 and 43:
3.2.2 Peer to Peer As we already me
Page 44 and 45:
5. Audio: by adding a MEMS micropho
Page 46 and 47:
Table 3.1: Material thicknesses and
Page 48 and 49:
(a) Node Front (b) Node Back Figure
Page 50 and 51:
Figure 3-9: Skin Patch mechanical t
Page 52 and 53:
used, which only further complicate
Page 54 and 55:
8. Teardrops and filleting were add
Page 56 and 57:
(a) 7.4V, 2.2mAh Li-ion Battery (b)
Page 58 and 59:
in more detail in the Results Chapt
Page 60 and 61:
has a 30mA forward current with a f
Page 62 and 63:
Figure 3-14: Strain gage layer prot
Page 64 and 65:
an unbalanced Wheatstone bridge, fr
Page 66 and 67:
Figure 3-16: QTC response curve for
Page 68 and 69:
(a) Prototype boards used to calibr
Page 70 and 71:
whisker is epoxied to the package o
Page 72 and 73:
Sound The only sensing modality add
Page 74 and 75:
Figure 3-20: Brain regular PCB fabr
Page 76 and 77:
3.4.3 Communications unit The Commu
Page 79 and 80:
Chapter 4 Software “A Scientist d
Page 81 and 82:
Networks. Why is this relevant? Mai
Page 83 and 84:
Figure 4-1: Bootloader memory map C
Page 85 and 86:
Bit Flag Name Description 0x01 FILE
Page 87 and 88: Figure 4-3: Sampling timing example
Page 89 and 90: We also mentioned before that there
Page 91 and 92: Code Instruction Description 0xA0 R
Page 93 and 94: Component Description 1 Indicates t
Page 95 and 96: last chapter. Figure 4-5: Automatic
Page 97 and 98: An important point to stress is tha
Page 99 and 100: (a) Rotation control (b) Zooming co
Page 101 and 102: 4.3 Communications Software The las
Page 103 and 104: 4.3.2 Peer to Peer Peer to peer com
Page 105 and 106: Chapter 5 Analysis of Results and E
Page 107 and 108: Figure 5-1: Stress concentration po
Page 109 and 110: of the assembled skin patch don’t
Page 111 and 112: Node SG1 SG2 R(X) R(Y) Δ(X) Δ(Y)
Page 113 and 114: Or, ∆R = R3 R3 R3 + RC (5.2) Howe
Page 115 and 116: R3 Microstrain 162Ω −99.7337969
Page 117 and 118: Figure 5-3: Strain Gage electronic
Page 119 and 120: (a) Data from both strain gages (b)
Page 121 and 122: (a) Pressure data being extracted f
Page 123 and 124: many of these did not work as expec
Page 125 and 126: Data plots were also extracted for
Page 127 and 128: (a) Temperature sensor data plots (
Page 129 and 130: Figure 5-12: Raw microphone data wi
Page 131 and 132: (a) Activity sensor data (b) Whiske
Page 133 and 134: 5.3 Performance Metrics We have pre
Page 135 and 136: (a) I 2 C data when a 1500Ωs resi
Page 137: Component # per node Current consum
Page 141 and 142: Chapter 6 Conclusions and Future Wo
Page 143 and 144: surface can be used to sense pressu
Page 145 and 146: Implement P2P Communications: curre
Page 147 and 148: Appendix A Abbreviations and Symbol
Page 149 and 150: Appendix B Bill of Materials 149
Page 151 and 152: Appendix C Schematics and PCB Layou
Page 153 and 154: 1 1 2 2 3 3 4 4 D D C C B B A A Tit
Page 155 and 156: Amp1 P0Amp101 P0Amp102 P0Amp103 P0A
Page 157 and 158: P0Cin02 P0East06 P0North06 P0South0
Page 159 and 160: Cain P0C ain01 P0Cain02 L1 P0L101 P
Page 161 and 162: Figure C-10: Node Bottom Overlay 16
Page 163 and 164: 4 3 2 1 A A Brain Communications Ac
Page 165 and 166: C3v P0C3v01 P0C3v02 Cavcc P0Cavcc01
Page 167 and 168: Appendix D Firmware 167
Page 169 and 170: C:\Documents and Settings\Gerardo\M
Page 183 and 184: Appendix E PC Code 183
Page 185 and 186: 1 c:\Documents and Settings\Gerardo
Page 187 and 188: 1 c:\Documents and Settings\Gerardo
Page 189 and 190:
3 c:\Documents and Settings\Gerardo
Page 191 and 192:
Page 193 and 194:
Page 195 and 196:
Page 197 and 198:
Page 199 and 200:
Page 201 and 202:
12 c:\Documents and Settings\Gerard
Page 203 and 204:
Page 205 and 206:
Page 207 and 208:
Bibliography [1] John C. Ansel. Ski
Page 209 and 210:
[33] Joseph Jacobson, B Comiskey, C
Page 211:
[68] Vilis Tutis. The physiology of
show all

S.N.A.K.E.: A Dynamically Reconfigurable Artificial Sensate Skin ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?