Final Documentation
DECO3100 Information Visualization Design Studio
Assignment 1 – Data Sculpture Documentation
Julia Tang 307 204 367
or PDF
Data Characteristics
I was researching aviation incidents and the causes of each Australian state. At first, I just looked at the accidents and deaths; however there were other aviation factors and concerns that people probably do not know about. For example, there are various incidents, not just terrorist attacks that contribute to aviation safety. What was the most interesting cause (in my opinion) was the number of bird strike incidents occurring at the airports. The feedback also highlighted that this perspective was very interesting.
Datasets
This table below (created by ManyEyes) shows the bird strikes at major airports in Australia.
Source data: "Table 17: Birdstrikes at major aerodromes, 2000 to 31 December 2008" Aviation Statistics 1 January 1998 to 31 December 2008; ATSB Transport Safety Report. http://www.atsb.gov.au/publications/2009/pdf/AviationStats.pdf
That graph above states the total amount of bird strikes but the total amount of planes departing and arriving is different in each state. Hence I used the rate of bird strikes in comparison to the number of inbound and outbound flights. It shows a clearer amount of the states being affected.
Bird strike rate per 10 000 total movements
2002 2003 2004 2005 2006
ACT 1.04 0.57 2.18 3.88 5.25
NSW 1.54 2.32 2.92 2.73 2.5
NT 8.41 8.19 11.59 13.47 7.92
QLD 2.74 4.38 5.4 7.11 3.95
SA 2.89 3.54 5.71 5.7 5.1
TAS 7.98 5.62 3.26 8.56 9.16
VIC 2.05 1.89 3.08 3.21 3.67
WA 1.52 3.29 4.35 3.64 4.11
I have used recent data from 2000 to 2008 because previously in Australia, reporting birdstrikes only happened when there was major damage to an aircraft.
Pattern
My data attributes were the number of bird strikes, the total number of incidents, and the number of accidents. The pattern I found out was no matter in which state, the number of aviation incidents has maintained a stable slow increase, however the number of bird strikes increased dramatically.
Here is my compiled final data:
Australian Government: Department of Transport and Regional Services Database
Number of Australian Incidents involving Bird strike by State and Year
2000 2001 2002 2003 2004 2005 2006 2007 2008
ACT 21 24 11 9 22 33 48 40 31
NSW 136 134 156 175 220 226 233 276 277
NT 73 82 126 104 124 160 116 125 125
QLD 193 214 236 317 375 460 417 446 507
SA 55 79 76 66 114 123 105 103 102
TAS 11 42 38 30 32 58 43 65 62
VIC 72 80 71 63 111 133 138 90 127
WA 81 81 83 97 118 112 162 152 130
Number of Australian Incidents in Total by State and Year
2000 2001 2002 2003 2004 2005 2006 2007 2008
ACT 108 88 78 66 93 122 149 99 128
NSW 1831 1735 1940 1535 1517 2091 2188 2519 2405
NT 391 378 443 423 442 420 446 508 497
QLD 1389 1267 1322 1078 1170 1572 1699 1572 1904
SA 334 366 346 281 365 408 556 738 518
TAS 76 115 105 97 118 153 135 170 179
VIC 849 756 797 582 590 806 969 843 892
WA 619 577 588 572 593 599 1031 1048 1017
Number of Australian Accidents involving Bird strike by State and Year
2000 2001 2002 2003 2004 2005 2006 2007 2008
ACT 0 0 0 0 0 0 0 0 0
NSW 1 3 1 0 0 2 0 0 1
NT 0 1 1 0 0 1 0 0 0
QLD 3 2 2 1 1 1 0 2 1
SA 0 0 0 0 0 1 0 0 0
TAS 0 0 0 0 0 0 0 0 0
VIC 0 0 0 0 0 0 0 1 0
WA 1 3 0 0 1 0 0 1 0
The accidents involving birds were more of an outlier; it didn’t actually mean anything because they only represented one or two accidents, not hundreds.
Now what is intriguing is the meaning behind the trend.
I researched a bit more, and there were some papers on the insight of why the number of bird strikes is increasing.
Modern aircrafts are increasing, with faster speed, and engines are quieter - increasing the probability as birds have less time to react. The surroundings of an airport have more grassland because residential property cannot be built do to the noise, so there are wetlands and agricultural lands which attract birds. Water is also more available at airfields, and wet weather can encourage the birds to feed on the grassland.
Source: “A Cost Effective Grassland Management Strategy to Reduce The Number of Birdstrikes.” http://eprints.qut.edu.au/16576/2/02whole.pdf
Design Rationale
The dimensions and form of the sculpture should be “driven by the data”. Hence, the physical form is actually the literal translation of these line graphs (the data derived from the table above).
The number of incidents involving bird strikes in Australian states:
To connect this graph and the one below was the data of the accidents involving bird strikes. There were only one to three accidents at most in each state, so the thickness of the joints characterized that number.
The total number of aviation incidents in Australian states:
The side view is supposed to look like this (minus the feathers), however, my actual model is not as smooth as I wanted it to be. It was because I used Maya 8.0 – 3D modelling software and was advised to use the 3D printer, hence polygons. I never really used polygons before and handling it was slightly different to Nurbs. I never expected it to be so ridged. At the end of the model, I smoothed it, but not enough.
Modelling
The length of each state is actually the most important, it is the bird strike rate per 10 000 total movements. With that, you can see which state is more affected as it compares the bird strikes to the number of aviation flights.
This is the NT sculpture part. What is seen is the back side with the bird strike data. What each plate has on it is the bird which caused the most number of incidents in the state.
The beginning for NSW, exploring the polygons.
The first encounter I had was putting an image plane in the background. Read a lot of tutorials to finally find the most simplest one. I created the NSW sculpture at home, and went to University computer labs to edit it. Unfortunately, the images did not turn out, I changed the location accordingly, but it would not show. So I edited the model to almost perfection. When I got home, I began receiving an error statement “Errors have occurred while reading this scene that may result in data loss. Please check the Script Editor for details”. I was really scared that I had lost all my work, and so went back to university to retrieve the files saved on the D drive. Met Eric along the way, and he told me it was just because of the Maya version I have is an old one. In the end, I redid the NSW data piece. Took me half the time of worrying.
I thought I had finished modelling, so ‘twisted’ the model. Then I read that labels had to be used so added on text. Painstakingly I added each letter, transformed it so it would fit the shape of the edge.
I added all the pieces together... Then I decided not to have the twist because it does not show the design as clear as it would normally. One deletion of the twisted line, it became normal, except for the text. Very annoying.
The pieces were all joined together roughly at the base. I hope the user would be able to interact it by looking at the different sides and through all the ‘gaps’ in between.
I would have ‘bent’ each state piece if there was just three. But I saw that there was time so I created more, I was able to create 5 states in total. It ended up being a fan shape, not entirely representing the propeller; it looks more like a bird to me.
I ‘smoothed’ the polygon, and it looks like this now:
Printing
The challenge here, was converting the Maya model to STL file for the 3D printer. I smoothed my design out, so there was over 900 000 facets (as you can see above compared to the simple polygon faces). I exported the Maya file to OBJ, but Rhinoceros could not open the OBJ file because it was 135Mb. It just kept on crashing to desktop.
I was so scared at this time that my design would not be able to be transferred across.
Had to ‘reduce’ the polygons in Maya (used the pc's in the Digital Hub - the computers there have like 8 Cores and 8Gb of RAM with really nice big screens). Unfortunately, it ended up not being so round and smoothed. Then imported the OBJ file into Rhino, took around 20 minutes to load. Saving as STL file was faster, but overall took an hour to finally import it to print.
Went downstairs to the cutting/3d printer/mill room, and Gabe helped me out with importing the file and fixing the triangular problems. I increased my scale by 150% because it just looked so small. Luckily I did it because much of the design was actually quite small. Sent to the 3D printer for $40.15.
Final Data Sculpture
Arrived the next morning to find Rick removing the sculptures from the 3D Printer. He then got rid of the excess powder, and started spraying water on it to place it in the oven.
The sculptures were to be baked for half an hour, and cooled for over half an hour.
Here is my sculpture, still warm from the oven:
The Q from QLD is missing.
As you can see here, some of the text did not come out. “Nankeen Kestrel” was half missing.
Otherwise, it’s a success.
I saw the other sculptures, and found that they were very aesthetic and curvy. I decided to paint the sculpture.
Painting
I found a box of acrylic paint (water based colours needed for this material).
I started mixing a little of the paints together (blue and yellow for green). But it was not enough, as this material just started absorbing everything so very fast. I had to do so many remixes I lost track of the tone.
Painting the entire sculpture took longer than I thought. Excluding the procrastination time, this took over 6 hours to paint because it was so fragile. Some of the joints in the middle crumbled a little on the side when painted.
It got a little messy because everything had to be fast paced or else everything would set.
The green represents the amount of wildlife conservation surrounding the airports. The darker the green, the more grassland there was.
The blue portrays the temperature, the darker the blue, the colder it is. I wanted to use red, but I thought it might turn out to look like a Christmas toy instead of the reality I wanted to represent.
I was moving it to paint it in a different direction, but Victoria broke off!
Really scared (again) because when I slightly touched the ends of the piece, it would just crumble. Notice the white powder on the newspaper, it felt like as if the whole model had turned to dust.
Using Tic Tac to hold the sculpture whilst the glue worked. Didn’t dare touch it until an hour later.
Started painting again the next day.
I did not want the green and blue to mix at the centre, because then that will distract the user from seeing the long joints (for the accidents). So I chose to colour it yellow, and the inside of the sculpture pale yellow.
Research and data sources
It was very difficult to find resources that originated from Australia and not United States, United Kingdom or European countries.
In the end, I found and used the data from the Australian Transport Safety Bureau. I found a lot of reports relating to aviation and incidents. What was great was that they had an online database of safety data. It was very concise, and that is what lead me to work on each state instead of Australia as a whole because each state had variable information.
Here are some sources I used for the final dataset.
NASA Aviation Safety Reporting System, http://www.37000feet.com/
ASRS Report “Bird or Animal Strike Reports”, http://asrs.arc.nasa.gov/docs/rpsts/animal.pdf
Aviation Accidents and Incidents, http://en.wikipedia.org/wiki/Aviation_accidents_and_incidents
Aircraft Accident Statistics, http://www.1001crash.com/index-page-statistique-lg-2.html
Accident Statistics, http://www.baaa-acro.com/Statistiques%20diverses.htm
Bureau of Transportation Statistics “Table 2-1: Transportation Fatalities by Mode“ http://www.bts.gov/publications/national_transportation_statistics/html/table_02_01.html
Plane Crash Info, http://www.planecrashinfo.com
Australian Transport Safety Bureau, Aviation Statistics Report, http://www.atsb.gov.au/publications/2009/AviationStats.aspx
Australian Transport Safety Bureau, Aviation Occurrence Database, http://statistics.dotars.gov.au/aviation/login.do?guest=guest&tableId=user/aviationuser/Accidents%20by%20State.txd
ATBS Aviation Research and Analysis Report, “Trends in immediately reportable matters involving regular public transport operations” Page 21 http://www.atsb.gov.au/publications/2007/pdf/b20070107_001.pdf
ATBS, “An Analysis of Australian Birdstrike Occurrences 2002-2006”, Page 36 http://www.atsb.gov.au/publications/2008/pdf/AR2008027.pdf
Bureau of Transport and Regional Economics, “Aviation Statistics” Page 17 http://www.bitre.gov.au/publications/54/Files/Digest2004-05.pdf
For more details involving the initial processes of the sculpture design, please read my blog at http://infostudio2009.ning.com/profile/Julia.
Assignment 1 – Data Sculpture Documentation
Julia Tang 307 204 367
or PDF
Data Characteristics
I was researching aviation incidents and the causes of each Australian state. At first, I just looked at the accidents and deaths; however there were other aviation factors and concerns that people probably do not know about. For example, there are various incidents, not just terrorist attacks that contribute to aviation safety. What was the most interesting cause (in my opinion) was the number of bird strike incidents occurring at the airports. The feedback also highlighted that this perspective was very interesting.
Datasets
This table below (created by ManyEyes) shows the bird strikes at major airports in Australia.
Source data: "Table 17: Birdstrikes at major aerodromes, 2000 to 31 December 2008" Aviation Statistics 1 January 1998 to 31 December 2008; ATSB Transport Safety Report. http://www.atsb.gov.au/publications/2009/pdf/AviationStats.pdf
That graph above states the total amount of bird strikes but the total amount of planes departing and arriving is different in each state. Hence I used the rate of bird strikes in comparison to the number of inbound and outbound flights. It shows a clearer amount of the states being affected.
Bird strike rate per 10 000 total movements
2002 2003 2004 2005 2006
ACT 1.04 0.57 2.18 3.88 5.25
NSW 1.54 2.32 2.92 2.73 2.5
NT 8.41 8.19 11.59 13.47 7.92
QLD 2.74 4.38 5.4 7.11 3.95
SA 2.89 3.54 5.71 5.7 5.1
TAS 7.98 5.62 3.26 8.56 9.16
VIC 2.05 1.89 3.08 3.21 3.67
WA 1.52 3.29 4.35 3.64 4.11
I have used recent data from 2000 to 2008 because previously in Australia, reporting birdstrikes only happened when there was major damage to an aircraft.
Pattern
My data attributes were the number of bird strikes, the total number of incidents, and the number of accidents. The pattern I found out was no matter in which state, the number of aviation incidents has maintained a stable slow increase, however the number of bird strikes increased dramatically.
Here is my compiled final data:
Australian Government: Department of Transport and Regional Services Database
Number of Australian Incidents involving Bird strike by State and Year
2000 2001 2002 2003 2004 2005 2006 2007 2008
ACT 21 24 11 9 22 33 48 40 31
NSW 136 134 156 175 220 226 233 276 277
NT 73 82 126 104 124 160 116 125 125
QLD 193 214 236 317 375 460 417 446 507
SA 55 79 76 66 114 123 105 103 102
TAS 11 42 38 30 32 58 43 65 62
VIC 72 80 71 63 111 133 138 90 127
WA 81 81 83 97 118 112 162 152 130
Number of Australian Incidents in Total by State and Year
2000 2001 2002 2003 2004 2005 2006 2007 2008
ACT 108 88 78 66 93 122 149 99 128
NSW 1831 1735 1940 1535 1517 2091 2188 2519 2405
NT 391 378 443 423 442 420 446 508 497
QLD 1389 1267 1322 1078 1170 1572 1699 1572 1904
SA 334 366 346 281 365 408 556 738 518
TAS 76 115 105 97 118 153 135 170 179
VIC 849 756 797 582 590 806 969 843 892
WA 619 577 588 572 593 599 1031 1048 1017
Number of Australian Accidents involving Bird strike by State and Year
2000 2001 2002 2003 2004 2005 2006 2007 2008
ACT 0 0 0 0 0 0 0 0 0
NSW 1 3 1 0 0 2 0 0 1
NT 0 1 1 0 0 1 0 0 0
QLD 3 2 2 1 1 1 0 2 1
SA 0 0 0 0 0 1 0 0 0
TAS 0 0 0 0 0 0 0 0 0
VIC 0 0 0 0 0 0 0 1 0
WA 1 3 0 0 1 0 0 1 0
The accidents involving birds were more of an outlier; it didn’t actually mean anything because they only represented one or two accidents, not hundreds.
Now what is intriguing is the meaning behind the trend.
I researched a bit more, and there were some papers on the insight of why the number of bird strikes is increasing.
Modern aircrafts are increasing, with faster speed, and engines are quieter - increasing the probability as birds have less time to react. The surroundings of an airport have more grassland because residential property cannot be built do to the noise, so there are wetlands and agricultural lands which attract birds. Water is also more available at airfields, and wet weather can encourage the birds to feed on the grassland.
Source: “A Cost Effective Grassland Management Strategy to Reduce The Number of Birdstrikes.” http://eprints.qut.edu.au/16576/2/02whole.pdf
Design Rationale
The dimensions and form of the sculpture should be “driven by the data”. Hence, the physical form is actually the literal translation of these line graphs (the data derived from the table above).
The number of incidents involving bird strikes in Australian states:
To connect this graph and the one below was the data of the accidents involving bird strikes. There were only one to three accidents at most in each state, so the thickness of the joints characterized that number.
The total number of aviation incidents in Australian states:
The side view is supposed to look like this (minus the feathers), however, my actual model is not as smooth as I wanted it to be. It was because I used Maya 8.0 – 3D modelling software and was advised to use the 3D printer, hence polygons. I never really used polygons before and handling it was slightly different to Nurbs. I never expected it to be so ridged. At the end of the model, I smoothed it, but not enough.
Modelling
The length of each state is actually the most important, it is the bird strike rate per 10 000 total movements. With that, you can see which state is more affected as it compares the bird strikes to the number of aviation flights.
This is the NT sculpture part. What is seen is the back side with the bird strike data. What each plate has on it is the bird which caused the most number of incidents in the state.
The beginning for NSW, exploring the polygons.
The first encounter I had was putting an image plane in the background. Read a lot of tutorials to finally find the most simplest one. I created the NSW sculpture at home, and went to University computer labs to edit it. Unfortunately, the images did not turn out, I changed the location accordingly, but it would not show. So I edited the model to almost perfection. When I got home, I began receiving an error statement “Errors have occurred while reading this scene that may result in data loss. Please check the Script Editor for details”. I was really scared that I had lost all my work, and so went back to university to retrieve the files saved on the D drive. Met Eric along the way, and he told me it was just because of the Maya version I have is an old one. In the end, I redid the NSW data piece. Took me half the time of worrying.
I thought I had finished modelling, so ‘twisted’ the model. Then I read that labels had to be used so added on text. Painstakingly I added each letter, transformed it so it would fit the shape of the edge.
I added all the pieces together... Then I decided not to have the twist because it does not show the design as clear as it would normally. One deletion of the twisted line, it became normal, except for the text. Very annoying.
The pieces were all joined together roughly at the base. I hope the user would be able to interact it by looking at the different sides and through all the ‘gaps’ in between.
I would have ‘bent’ each state piece if there was just three. But I saw that there was time so I created more, I was able to create 5 states in total. It ended up being a fan shape, not entirely representing the propeller; it looks more like a bird to me.
I ‘smoothed’ the polygon, and it looks like this now:
Printing
The challenge here, was converting the Maya model to STL file for the 3D printer. I smoothed my design out, so there was over 900 000 facets (as you can see above compared to the simple polygon faces). I exported the Maya file to OBJ, but Rhinoceros could not open the OBJ file because it was 135Mb. It just kept on crashing to desktop.
I was so scared at this time that my design would not be able to be transferred across.
Had to ‘reduce’ the polygons in Maya (used the pc's in the Digital Hub - the computers there have like 8 Cores and 8Gb of RAM with really nice big screens). Unfortunately, it ended up not being so round and smoothed. Then imported the OBJ file into Rhino, took around 20 minutes to load. Saving as STL file was faster, but overall took an hour to finally import it to print.
Went downstairs to the cutting/3d printer/mill room, and Gabe helped me out with importing the file and fixing the triangular problems. I increased my scale by 150% because it just looked so small. Luckily I did it because much of the design was actually quite small. Sent to the 3D printer for $40.15.
Final Data Sculpture
Arrived the next morning to find Rick removing the sculptures from the 3D Printer. He then got rid of the excess powder, and started spraying water on it to place it in the oven.
The sculptures were to be baked for half an hour, and cooled for over half an hour.
Here is my sculpture, still warm from the oven:
The Q from QLD is missing.
As you can see here, some of the text did not come out. “Nankeen Kestrel” was half missing.
Otherwise, it’s a success.
I saw the other sculptures, and found that they were very aesthetic and curvy. I decided to paint the sculpture.
Painting
I found a box of acrylic paint (water based colours needed for this material).
I started mixing a little of the paints together (blue and yellow for green). But it was not enough, as this material just started absorbing everything so very fast. I had to do so many remixes I lost track of the tone.
Painting the entire sculpture took longer than I thought. Excluding the procrastination time, this took over 6 hours to paint because it was so fragile. Some of the joints in the middle crumbled a little on the side when painted.
It got a little messy because everything had to be fast paced or else everything would set.
The green represents the amount of wildlife conservation surrounding the airports. The darker the green, the more grassland there was.
The blue portrays the temperature, the darker the blue, the colder it is. I wanted to use red, but I thought it might turn out to look like a Christmas toy instead of the reality I wanted to represent.
I was moving it to paint it in a different direction, but Victoria broke off!
Really scared (again) because when I slightly touched the ends of the piece, it would just crumble. Notice the white powder on the newspaper, it felt like as if the whole model had turned to dust.
Using Tic Tac to hold the sculpture whilst the glue worked. Didn’t dare touch it until an hour later.
Started painting again the next day.
I did not want the green and blue to mix at the centre, because then that will distract the user from seeing the long joints (for the accidents). So I chose to colour it yellow, and the inside of the sculpture pale yellow.
Research and data sources
It was very difficult to find resources that originated from Australia and not United States, United Kingdom or European countries.
In the end, I found and used the data from the Australian Transport Safety Bureau. I found a lot of reports relating to aviation and incidents. What was great was that they had an online database of safety data. It was very concise, and that is what lead me to work on each state instead of Australia as a whole because each state had variable information.
Here are some sources I used for the final dataset.
NASA Aviation Safety Reporting System, http://www.37000feet.com/
ASRS Report “Bird or Animal Strike Reports”, http://asrs.arc.nasa.gov/docs/rpsts/animal.pdf
Aviation Accidents and Incidents, http://en.wikipedia.org/wiki/Aviation_accidents_and_incidents
Aircraft Accident Statistics, http://www.1001crash.com/index-page-statistique-lg-2.html
Accident Statistics, http://www.baaa-acro.com/Statistiques%20diverses.htm
Bureau of Transportation Statistics “Table 2-1: Transportation Fatalities by Mode“ http://www.bts.gov/publications/national_transportation_statistics/html/table_02_01.html
Plane Crash Info, http://www.planecrashinfo.com
Australian Transport Safety Bureau, Aviation Statistics Report, http://www.atsb.gov.au/publications/2009/AviationStats.aspx
Australian Transport Safety Bureau, Aviation Occurrence Database, http://statistics.dotars.gov.au/aviation/login.do?guest=guest&tableId=user/aviationuser/Accidents%20by%20State.txd
ATBS Aviation Research and Analysis Report, “Trends in immediately reportable matters involving regular public transport operations” Page 21 http://www.atsb.gov.au/publications/2007/pdf/b20070107_001.pdf
ATBS, “An Analysis of Australian Birdstrike Occurrences 2002-2006”, Page 36 http://www.atsb.gov.au/publications/2008/pdf/AR2008027.pdf
Bureau of Transport and Regional Economics, “Aviation Statistics” Page 17 http://www.bitre.gov.au/publications/54/Files/Digest2004-05.pdf
For more details involving the initial processes of the sculpture design, please read my blog at http://infostudio2009.ning.com/profile/Julia.
Add a Comment
About
© 2009 Created by Andrew on Ning. Create a Ning Network!
You need to be a member of infostudio 2009 to add comments!
Join this Ning Network