Abstract image of a woman sitting on books and studying on a laptop

Making The Invisible Visible: Visualising Air Quality In An Understandable Way

Ye Lin, profile picture
Ye Lin

The document explores how to effectively visualize air quality data from fixed sensors in a way that is engaging and understandable for non-scientists, specifically in a mobile context. Through user interviews and experiments, it identifies the need for clearer and more playful visualizations to encourage public awareness and usage of air quality information. Several design recommendations and concepts, including the use of photographs and interactive features, are proposed to make air quality data more relatable and promote proactive behaviors among users.

Technology
Related topics:
Air Quality Insights • Mobile App Development
1 of 64
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
Making The Invisible Visible Visualising Air Quality In An Understandable Way Ye Lin Supervisors: Lorna Wall, Han Pham | 09/2014
RESEARCH QUESTION “How can air quality data from fixed location sensors be visualised in an understandable and playful way to users in a mobile context?”
PROCESS MARCH - MAY JUNE JULY AUGUST SEPTEMBER Design Iterations 6 London Residents & 1 Beijing Resident 8 London Residents User Interviews User Experiments iOS app & website Development Literature Review Future Works
PROBLEM SOURCE: Air quality information on London Evening Standard newspaper Scientific Data Non-scientist In an understandable and engaging format VISUALISATION
AIR POLLUTION The presence of contaminants or pollutant substances in the air that interfere with human health or welfare, or produce other harmful environmental effects. n. United States Environmental Protection Agency (2007) 1 Nitrogen dioxide (NO2) is less visible. ! The forming of air pollution is an accumulative process. ! The health effects are hard to be linked to specific human behaviours. 2 3 Scientific Data Non-scientist In an understandable and engaging format VISUALISATION
1.Map Based [1] [2] [1] London Air: http://www.londonair.org.uk/london/asp/PublicEpisodes.asp [2] Defra: http://uk-air.defra.gov.uk/latest/ [3] London Air App: http://itunes.apple.com/gb/app/london-air/id358970517?mt=8 [3]
[2] 1 [3] Difficult to show temporal variations 1.Map Based [1] 1 Good to show spatial variations Lack motivation to check the data 2 Explanation text to visualisations Available on web and mobile 2 Hard to read [1] London Air: http://www.londonair.org.uk/london/asp/PublicEpisodes.asp [2] Defra: http://uk-air.defra.gov.uk/latest/ [3] London Air App: http://itunes.apple.com/gb/app/london-air/id358970517?mt=8
2.Numeric Data [1] [2] [3] [1] The London Marylebone Road AQI: http://aqicn.org/city/united-kingdom/london-marylebone-road/m/hk/&aboutaqi [2] China Air Quality Index App: https://itunes.apple.com/us/app/china-air-quality-index/id477700080?mt=8 [3] China Air App: https://itunes.apple.com/us/app/chinas-air/id777458271?mt=8
2.Numeric Data [1] [2] [3] 1 Difficult to show spatial & temporal variations 1 Good to show current air quality [1] The London Marylebone Road AQI: http://aqicn.org/city/united-kingdom/london-marylebone-road/m/hk/&aboutaqi [2] China Air Quality Index App: https://itunes.apple.com/us/app/china-air-quality-index/id477700080?mt=8 [3] China Air App: https://itunes.apple.com/us/app/chinas-air/id777458271?mt=8 Combined with other kinds of visualisations 2 Use numeric data, colour blocks, health advice 2 People don’t understand “moderate” air quality Use photographs to add an emotional connection to data Use healthy, unhealthy, understandable words
3.Line Graph [1] 1 InAir [1]! ! • People were interested to see spikes in visualisations. ! • Give users data and let them decide what to do Difficult to show spatial variations 1 Good to show temporal variations [1] Kim, S., Paulos, E. & Mankoff, J. (2013). inAir: A Longitudinal study of Indoor Air Quality Measurements and Visualizations. CHI 2013, April 27 – May 2, 2013, Paris, France.
Public Usage People are not aware of the existing air quality information Personal! Usage Channels: web, newspaper, mobile app ! Format: text, map, numeric data and line graph. 2 Designed based on the low spatial and/or temporal resolution data set in cities 1 EXISTING VISUALISATIONS
Public Usage Personal! Usage Channels: web, newspaper, mobile app ! Format: text, map, numeric data and line graph. People are not aware of the existing air quality information 2 Designed based on the low spatial and/or temporal resolution data set in cities 1 Not well applied to ! a mobile context 3 EXISTING VISUALISATIONS
No personalised data 1 Didn’t show temporal variations 2 Didn’t address on the NO2 pollution 3 Depressing to check data 4 5 Not playful EXISTING MOBILE VISUALISATIONS [1] [2] [3] [1] AQ Scotland App: https://itunes.apple.com/en/app/air-quality-in-scotland/id838197830 [2] City Air App: https://itunes.apple.com/gb/app/city-air/id706049131?mt=8 [3] Birkett Index App: https://itunes.apple.com/gb/app/birkett-index/id646281816?mt=8
AIR QUALITY INDEX • AQI is widely used in air quality visualisations • Inconsistent between countries • No standard threshold to define the healthy and unhealthy levels. No thresholds of health effects can be identified (Laden et al, 2006)! Because people’s susceptibility to pollution levels varies, and health effects on different people also varies. The UK AQI: SOURCE: Laden, F., Schwartz, J., Speizer, F. E., & Dockery, D. W. (2006). Reduction in fine particulate air pollution and mortality - Extended follow-up of the Harvard six cities study. American Journal of Respiratory and Critical Care Medicine, 173(6), 667-672. doi: Doi 10.1164/Rccm.200503-443oc
Policy-defined NO2 limits for cities. ! No standard for citizen exposure is available. EU Annual Mean Limit (Healthy for general population) 40 μg/m3 EU Hourly Mean Limit UK AQI Low Band 200 μg/m3 (Healthy for at-risk group) UK AQI Moderate Band 200 μg/m3 400 μg/m3 (Not to be exceeded 18 times per year) COMEAP, Committee on the Medical Effects of Air Pollutants (2011). Review of the UK Air Quality Index, A report by the Committee on the Medical Effects of Air Pollutants. Retrieved from http://www.hpa.org.uk/webc/HPAwebFile/HPAweb_C/1317137023144 Air Quality Standards from European Commission. Retrieved from http://ec.europa.eu/ environment/air/quality/standards.htm
Making The Invisible Visible: Visualising Air Quality In An Understandable Way
People’s susceptibility to pollution levels varies! ! Volumes of pollutants vary temporally. 1 2
! It is difficult to decide on a healthy/unhealthy volume.
! It is difficult to decide the measured length of the averaging period.
VISUALISATIONS PLAYFUL WearAir Air quality information was visualised not only to the wearer but also to the people around, which shows the importance of making air quality visualisation playful and engaging in order to raise the users’ initiative to share the visualisation with others. SOURCE: Kim, S., Paulos, E. & Gross, M. (2010). WearAir: Expressive T-shirts for Air Quality Sensing, TEI 2010, January 25–27, 2010, Cambridge, Massachusetts, USA
Balloons Participants were encouraged to take balloons with them to different parts of the city to explore air quality by themselves. They were interested and surprised to see changes of air quality and came up with ideas on why the air quality was changing regarding to their behaviours. VISUALISATIONS PLAYFUL SOURCE: Kuznetsov, S., Davis, G., Paulos, E., Gross, M. and Cheung, J. (2011). Red Balloon, Green Balloon, Sensors in the Sky. UbiComp 11, September 17–21, 2011, Beijing, China.
Floating Beijing SOURCE:Guler, D. & Wang, X. (2012). FLOAT Beijing. Retrieved from http://f-l-o-a-t.com/ Because people were finding it hard to see the stars in the Beijing night sky, they visualised things that people should be able to see but could not see due to pollution, and provided a strong visual contrast. VISUALISATIONS PLAYFUL
[1] [2] [3] • Simple visualisations ! ! • Visual contrast! ! • Playful and engaging! ! • Show changes of AQ data VISUALISATIONS PLAYFUL [1] Kim, S., Paulos, E. & Gross, M. (2010). WearAir: Expressive T-shirts for Air Quality Sensing, TEI 2010, January 25–27, 2010, Cambridge, Massachusetts, USA [2] Kuznetsov, S., Davis, G., Paulos, E., Gross, M. and Cheung, J. (2011). Red Balloon, Green Balloon, Sensors in the Sky. UbiComp 11, September 17–21, 2011, Beijing, China. [3] Guler, D. & Wang, X. (2012). FLOAT Beijing. Retrieved from http://f-l-o-a-t.com/
[1] [2] [3] • Data from mobile sensors! ! • Not quality data VISUALISATIONS PLAYFUL [1] Kim, S., Paulos, E. & Gross, M. (2010). WearAir: Expressive T-shirts for Air Quality Sensing, TEI 2010, January 25–27, 2010, Cambridge, Massachusetts, USA [2] Kuznetsov, S., Davis, G., Paulos, E., Gross, M. and Cheung, J. (2011). Red Balloon, Green Balloon, Sensors in the Sky. UbiComp 11, September 17–21, 2011, Beijing, China. [3] Guler, D. & Wang, X. (2012). FLOAT Beijing. Retrieved from http://f-l-o-a-t.com/
SENSORS FIXED LOCATION LAQN Data Set Regulatory grade sensors Sensing London Project Sensor Cost High cost Lower cost Data Quality High quality & calibrated High quality & calibrated Spatial Resolution of Data Low spatial resolution High spatial resolution Temporal Resolution of Data Low temporal resolution High temporal resolution LAQN: http://www.londonair.org.uk/LondonAir/Default.aspx Sensing London project: https://futurecities.catapult.org.uk/project-full-view/-/asset_publisher/oDS9tiXrD0wi/content/project-sensing-london/
Reach a wider audience PEOPLE Show changes of ! air quality data 4 Give people air quality data, and let them decide what to do VISUALISATIONS PLAYFUL Visual objects, or things that people could not see because of pollution Simple & understandable ways Interactive lights, and by changing colours (red, yellow and green), volumes or by blinking 1 2 Playful and engaging ways Not only to the user but to people around 3 Visual contrast
Visual objects, or things that people could not see because of pollution Simple & understandable ways Interactive lights, and by changing colours (red, yellow and green), volumes or by blinking 1 2 Playful and engaging ways not only to the user but to people around 3 Visual contrast Show changes of air quality data 4 Give people air quality data, and let them decide what to do Fit into people’s daily lives 5 Use habitual behaviours VISUALISATIONS REQUIREMENTS
PERCEIVED INDICATORS Air Quality! Understandable PERSONAL! EXPERIENCE NOTICE POLLUTION Visual cues, odours, residue, health effects Trigger interests / read more articles / come up with solutions to avoid or reduce pollution USER INTERVIEWS FINDINGS
Participants understood the air quality was bad via visual cues. Photographs
[2] [3] [1] [1] The Atlantic (2012). A Stunning Visualization of China's Air Pollution. Retrieved from: http:// www.theatlantic.com/ international/archive/2012/07/ a-stunning-visualization-of-chinas- air-pollution/259455/ [2] China Air Daily: http:// www.chinaairdaily.com/ [3] Guler, D. (2012). FloatPM. Interactive Art and Computational Design. Retrieved from http:// golancourses.net/2012spring/ 02/09/deren-guler_ project1_float-pm/ VISUALISATION PHOTOGRAPHS
[2] [3] [1] [1] The Atlantic (2012). A Stunning Visualization of China's Air Pollution. Retrieved from: http://www.theatlantic.com/international/archive/2012/07/ a-stunning-visualization-of-chinas-air-pollution/259455/ [2] China Air Daily: http://www.chinaairdaily.com/ [3] Guler, D. (2012). FloatPM. Interactive Art and Computational Design. Retrieved from http://golancourses.net/2012spring/02/09/deren-guler_ project1_float-pm/ 1 Used original photos Addressed on the PM pollution in Beijing 2 3 No data over photographs 1 Comparison is a key to help people understand 1) To understand if the current data 2) To understand the trend of air quality But NO2 is less visible… VISUALISATION PHOTOGRAPHS
“When I saw the number I can feel how good or bad the air quality level will be for today, because I’m used to it.” (P7) — A Beijing resident, using the app for two years Interpret the level of severity of air pollution to the experience in real life from reading the AQI number. USER INTERVIEWS FINDINGS China Air Quality Index App: https://itunes.apple.com/us/app/china-air-quality-index/id477700080?mt=8
Make the invisible visible Display data + DESIGN CONCEPT
Participants want to see… Temporal! Variations Changes & trends! in past 10 years? Spatial! Variations ! at my home? ! at my work place?! in Beijing? How is the air ! quality at my ! current location? Health effects? The safe/dangerous level? Actions can take? Causes?
DESIGN INSTANO2
CAMERA! VIEW Take a photo APPLY! FILTERS SHARE! THE PHOTO MY PHOTO! STREAM INSTANO2 MODEL
HEALTHY AND UNHEALTHY THRESHOLD The EU annual mean value of 40 μg/m3 ! INFORMATION: Numeric data, colour block, a healthy/unhealthy label the user’s current location and time Only NO2 data INSTANO2 VISUALISATION
Origin! photo INSTANO2 VISUALISATION Make the invisible visible via photographs Adjust the appearance of the photograph based on the air quality data
INSTANO2 CAMERA VIEW Take a photo to explore air quality The current NO2 level at the current location
INSTANO2 LOCATION FILTER Compare the current air quality data with OTHER PLACES: Places that people are familiar with, such as places in London, major cities in the UK, capital cities in Europe and over the world.
INSTANO2 TIME FILTER Compare the current air quality data with HISTORICAL DATES: Yesterday, last week, last month, the same day in past months of this year, and the same date in past years.
INSTANO2 SHARING 1 Talk and share air quality data 2 3 Social media Photos collection website My Places
Real time air quality data at places,! home, work places, places people passing by everyday. Motivate people to check air quality Like checking weather at places INSTANO2 MY PLACES
Experimental area, LAQN data set, low spatial and temporal resolution Take six pictures using the app Participants were encouraged to go anywhere they liked to and the task was to take six pictures using the app. USER EXPERIMENTS DESIGN
USER EXPERIMENTS FINDINGS “ Places participants went to Main streets! Took photos with lots of traffic, especially with buses, expecting to see poor air quality. Places they went to everyday! Gyms where they swam, streets they frequently walked on, shops where they grabbed sandwiches, cafes where they had lunch and parks where they relaxed. To understand the air quality they were exposed to every day. Places they’ve never been to! Cafes and hospitals nearby, to check whether they should go there in the future, to explore good spots to go to. 1 2 3
USER EXPERIMENTS FINDINGS “ Places where they would like to use the app 1 2 3 Home and work/study places! Check air quality frequently ! Places with lots of traffic and people! Oxford Street and Regent Street. ! Tourist places! River Thames and Kew Garden. These were places where they would usually take pictures and by using the app, they could also check if the air quality was good or bad there.
USER EXPERIMENTS FINDINGS “People like taking photos. It’s easy to do.” (E3) “I like the camera interface. Users don’t have to learn a new interface.” (E4) The app was easy to use. Participants were using the app just like using other camera apps.
USER EXPERIMENTS FINDINGS “I think colour is enough to represent the current air quality. But photos made it more real, and give you a sense of how good or bad it might affect you. When I see the current picture applied with a poor air quality filter, I feel dirty and a little uncomfortable, and I want to move away from that place soon.” (E5) Photographs visualised the effects of air pollution that could affect him.
USER EXPERIMENTS FINDINGS
USER EXPERIMENTS FINDINGS
Location filters Understand the spatial variations of air quality • Search function • Photo collage USER EXPERIMENTS FINDINGS
Time filters “I noticed the air quality is far better in the winter.” (E7) • Understand temporal variations of air quality • Reflect in humorous ways USER EXPERIMENTS FINDINGS
Pre-study interviews During Experiments Depressing to check data Participants reflected in humorous ways! Have no idea how air quality data looks like even when they were facing poor air quality. The app sparked curiosity, triggered exploration and raised awareness Participants thought and talked about what air quality meant, where the pollution came from, what might change the air quality and why it was changing. They also asked for more explanations about the data, expressed their own initiative to interpret and explore the data further to learn more. USER EXPERIMENTS FINDINGS
Data is lacking regarding physical locations and not dynamic. Experimental area, LAQN data set, low spatial and temporal resolution Difficult to show changes DESIGN CHALLENGES
DESIGN CHALLENGES SENSOR 1 SENSOR 2 27 110 CHANGES 1:! Participants can see changes of data when walk in between sensors! Data source changes from one sensor to the other.
CHANGES 2: On camera view, show a more dynamic data by adding a random variation on data. Data from the nearest sensor Data varies by a range of 10 by every 4 seconds X [X, X+10] 100 [100, 110] 55 [55, 65] DESIGN CHALLENGES
USER EXPERIMENTS FINDINGS One participant lost interest using the app Need a higher spatially dense data set DESIGN OPPORTUNITY Request Sensors! The app says data not available, or that the nearest sensor is 5 miles away, the user could press a button and request city authorities to deploy sensors in that area. DESIGN OPPORTUNITY Use computer vision! ! Recognise the content in the photo, and adjust NO2 levels based on numbers of cars, buses, etc.. Because he noticed the low spatially dense data set, changes in data didn’t match with his expectations
Public Usage Personal Usage More customisation PUSH • Search function for locations, make photo collages using the app. • Highlight key event dates in the time filters, such as the introduction of the congestion charge in London. More spatially dense data set INSTANO2 FUTURE WORKS Air Quality Data from fixed location sensors Users • More rational spikes can be seen from the data. A longer experiment • to learn about their motivations to check air quality data and their behaviour changes. 1 2 3
Public Usage Personal Usage GENERATE INSTANO2 FUTURE WORKS 1 User usage data (time/location data) Air Quality Data from fixed location sensors PUSH Users 2 Photo-based data USE Other stakeholders, such as city officials
Public Usage Personal Usage 1 User usage data (time/location data) Data to city officials Where Where citizens are interested to see air quality data When When people are using the app Who Who these users are Locations where air quality sensors are needed. Timing when air quality data needs to be pushed to the users.. Target Group to whom air quality is important. INSTANO2 FUTURE WORKS
Public Usage Personal Usage INSTANO2 FUTURE WORKS 2 Photo-based data add value to numerical data from the sensors Photos Photos collection • #searchable • sharable • contextual • Visualisation of London’s air quality • Crowdsourced insights from citizens • To learn citizens’ conversations about air quality
Making The Invisible Visible: Visualising Air Quality In An Understandable Way
Visualisation of London’s air quality! Crowdsourced insights from citizens To learn citizens’ conversations about air quality
Thanks! Q&A? Ye Lin Supervisors: Lorna Wall, Han Pham | 09/2014

More Related Content

DOC
BREATHE Project
Nijaz Deleut Kemo
 
PDF
33 examples of Data art and data installation
Mireille Raad
 
PPTX
IOT ENABLED EMBEDDED BASED REAL TIME AIR QUALITY MONITORING SYSTEM
MOHAMMED FURQHAN
 
PDF
Air quality and urban design
Gustavo Olivares
 
PPTX
Group-8-Project
Rahul Kumar Thakur
 
PPT
Citizen Sensing Efforts November 2011
timsdye
 
PDF
IRJET - Air Quality Index – A Study to Assess the Air Quality
IRJET Journal
 
PDF
IRJET- IoT based Air Pollution Monitoring System to Create a Smart Environment
IRJET Journal
 
BREATHE Project
Nijaz Deleut Kemo
 
33 examples of Data art and data installation
Mireille Raad
 
IOT ENABLED EMBEDDED BASED REAL TIME AIR QUALITY MONITORING SYSTEM
MOHAMMED FURQHAN
 
Air quality and urban design
Gustavo Olivares
 
Group-8-Project
Rahul Kumar Thakur
 
Citizen Sensing Efforts November 2011
timsdye
 
IRJET - Air Quality Index – A Study to Assess the Air Quality
IRJET Journal
 
IRJET- IoT based Air Pollution Monitoring System to Create a Smart Environment
IRJET Journal
 

Similar to Making The Invisible Visible: Visualising Air Quality In An Understandable Way (20)

PDF
Dt24761766
IJERA Editor
 
PDF
Air Quality Visualization
IRJET Journal
 
PPTX
Air Quality Study on measeurement tools.pptx
RAJA SEKHAR MAMILLAPALLI
 
PDF
HOSTILE GAS MONITORING SYSTEM USING IoT
IRJET Journal
 
PDF
A Literature Review on Ambient Air Quality Monitoring Near the Solid Waste Di...
IRJET Journal
 
PDF
Group-8-Summary
Rahul Kumar Thakur
 
PDF
Pollution Monitoring System using Arduino and various gas sensor
Utkarsh Jaiswal
 
PDF
Air Monitoring in Detail
Manik Bhola
 
PDF
Design and Implementation of Portable Outdoor Air Quality Measurement System ...
IJECEIAES
 
PDF
SIMULATION OF ATMOSPHERIC POLLUTANTS DISPERSION IN AN URBAN ENVIRONMENT
AM Publications
 
PDF
AIR POLLUTION IMPACT ON ENVIRONMENT IN KOZHIKODE CITY
IRJET Journal
 
PDF
AIR POLLUTION IMPACT ON ENVIRONMENT IN KOZHIKODE CITY
IRJET Journal
 
PDF
EDD Project A35 group. final.pdf Department of ENTC
MihirDatir1
 
PDF
IRJET- Air Pollution Monitoring System using the Internet of Things
IRJET Journal
 
PDF
Air Quality Monitoring and Control System in IoT
ijtsrd
 
PDF
AIR QUALITY INDEX MEASUREMENT
Dhruvan Delvadia
 
PDF
19_Bigdasdsadsadsadasdasddsadsd Data.pdf
NV5NV5
 
PDF
IRJET- IoT based Industrial Pollution Monitoring
IRJET Journal
 
PDF
Air Pollution Prediction via Differential Evolution Strategies with Random Fo...
IRJET Journal
 
PDF
IRJET- Smart Water Meter Reading and Distribution
IRJET Journal
 
Dt24761766
IJERA Editor
 
Air Quality Visualization
IRJET Journal
 
Air Quality Study on measeurement tools.pptx
RAJA SEKHAR MAMILLAPALLI
 
HOSTILE GAS MONITORING SYSTEM USING IoT
IRJET Journal
 
A Literature Review on Ambient Air Quality Monitoring Near the Solid Waste Di...
IRJET Journal
 
Group-8-Summary
Rahul Kumar Thakur
 
Pollution Monitoring System using Arduino and various gas sensor
Utkarsh Jaiswal
 
Air Monitoring in Detail
Manik Bhola
 
Design and Implementation of Portable Outdoor Air Quality Measurement System ...
IJECEIAES
 
SIMULATION OF ATMOSPHERIC POLLUTANTS DISPERSION IN AN URBAN ENVIRONMENT
AM Publications
 
AIR POLLUTION IMPACT ON ENVIRONMENT IN KOZHIKODE CITY
IRJET Journal
 
AIR POLLUTION IMPACT ON ENVIRONMENT IN KOZHIKODE CITY
IRJET Journal
 
EDD Project A35 group. final.pdf Department of ENTC
MihirDatir1
 
IRJET- Air Pollution Monitoring System using the Internet of Things
IRJET Journal
 
Air Quality Monitoring and Control System in IoT
ijtsrd
 
AIR QUALITY INDEX MEASUREMENT
Dhruvan Delvadia
 
19_Bigdasdsadsadsadasdasddsadsd Data.pdf
NV5NV5
 
IRJET- IoT based Industrial Pollution Monitoring
IRJET Journal
 
Air Pollution Prediction via Differential Evolution Strategies with Random Fo...
IRJET Journal
 
IRJET- Smart Water Meter Reading and Distribution
IRJET Journal
 
Ad

Recently uploaded (20)

PDF
Getting Started with Data Integration: FME Form 101
Safe Software
 
PDF
A Late Bloomer's Guide to GenAI: Ethics, Bias, and Effective Prompting - Boha...
Mindy Bohannon
 
PDF
How ambidextrous entrepreneurial leaders react to the artificial intelligence...
IAESIJAI
 
PPTX
O2C Customer Invoices to Receipt V15A.pptx
ssuserbfa915
 
DOCX
search engine optimization ppt fir known well about this
StandardCodeLearning
 
PDF
A review of recent deep learning applications in wood surface defect identifi...
IAESIJAI
 
PDF
NewMind AI Weekly Chronicles – August ’25 Week III
NewMind AI
 
PDF
Zenith AI: Advanced Artificial Intelligence
Biz Preneurs
 
PDF
Architecture types and enterprise applications.pdf
ChristopherTHyatt
 
PDF
Hybrid model detection and classification of lung cancer
IAESIJAI
 
PPTX
Group 1 Presentation -Planning and Decision Making .pptx
MatthewLewis227954
 
PDF
Five Habits of High-Impact Board Members
OnBoard
 
PDF
Video forgery: An extensive analysis of inter-and intra-frame manipulation al...
IAESIJAI
 
PDF
Hybrid horned lizard optimization algorithm-aquila optimizer for DC motor
IAESIJAI
 
PDF
Enhancing emotion recognition model for a student engagement use case through...
IAESIJAI
 
PPTX
Web Crawler for Trend Tracking Gen Z Insights.pptx
Actowiz Solustions
 
PPTX
Benefits of Physical activity for teenagers.pptx
Nokwanda Thabethe
 
PPTX
MicrosoftCybserSecurityReferenceArchitecture-April-2025.pptx
SilvioHayashi
 
PDF
WOOl fibre morphology and structure.pdf for textiles
Rajendrakumar868651
 
PDF
A comparative study of natural language inference in Swahili using monolingua...
IAESIJAI
 
Getting Started with Data Integration: FME Form 101
Safe Software
 
A Late Bloomer's Guide to GenAI: Ethics, Bias, and Effective Prompting - Boha...
Mindy Bohannon
 
How ambidextrous entrepreneurial leaders react to the artificial intelligence...
IAESIJAI
 
O2C Customer Invoices to Receipt V15A.pptx
ssuserbfa915
 
search engine optimization ppt fir known well about this
StandardCodeLearning
 
A review of recent deep learning applications in wood surface defect identifi...
IAESIJAI
 
NewMind AI Weekly Chronicles – August ’25 Week III
NewMind AI
 
Zenith AI: Advanced Artificial Intelligence
Biz Preneurs
 
Architecture types and enterprise applications.pdf
ChristopherTHyatt
 
Hybrid model detection and classification of lung cancer
IAESIJAI
 
Group 1 Presentation -Planning and Decision Making .pptx
MatthewLewis227954
 
Five Habits of High-Impact Board Members
OnBoard
 
Video forgery: An extensive analysis of inter-and intra-frame manipulation al...
IAESIJAI
 
Hybrid horned lizard optimization algorithm-aquila optimizer for DC motor
IAESIJAI
 
Enhancing emotion recognition model for a student engagement use case through...
IAESIJAI
 
Web Crawler for Trend Tracking Gen Z Insights.pptx
Actowiz Solustions
 
Benefits of Physical activity for teenagers.pptx
Nokwanda Thabethe
 
MicrosoftCybserSecurityReferenceArchitecture-April-2025.pptx
SilvioHayashi
 
WOOl fibre morphology and structure.pdf for textiles
Rajendrakumar868651
 
A comparative study of natural language inference in Swahili using monolingua...
IAESIJAI
 
Ad

Making The Invisible Visible: Visualising Air Quality In An Understandable Way

  • 1. Making The Invisible Visible Visualising Air Quality In An Understandable Way Ye Lin Supervisors: Lorna Wall, Han Pham | 09/2014
  • 2. RESEARCH QUESTION “How can air quality data from fixed location sensors be visualised in an understandable and playful way to users in a mobile context?”
  • 3. PROCESS MARCH - MAY JUNE JULY AUGUST SEPTEMBER Design Iterations 6 London Residents & 1 Beijing Resident 8 London Residents User Interviews User Experiments iOS app & website Development Literature Review Future Works
  • 4. PROBLEM SOURCE: Air quality information on London Evening Standard newspaper Scientific Data Non-scientist In an understandable and engaging format VISUALISATION
  • 5. AIR POLLUTION The presence of contaminants or pollutant substances in the air that interfere with human health or welfare, or produce other harmful environmental effects. n. United States Environmental Protection Agency (2007) 1 Nitrogen dioxide (NO2) is less visible. ! The forming of air pollution is an accumulative process. ! The health effects are hard to be linked to specific human behaviours. 2 3 Scientific Data Non-scientist In an understandable and engaging format VISUALISATION
  • 6. 1.Map Based [1] [2] [1] London Air: http://www.londonair.org.uk/london/asp/PublicEpisodes.asp [2] Defra: http://uk-air.defra.gov.uk/latest/ [3] London Air App: http://itunes.apple.com/gb/app/london-air/id358970517?mt=8 [3]
  • 7. [2] 1 [3] Difficult to show temporal variations 1.Map Based [1] 1 Good to show spatial variations Lack motivation to check the data 2 Explanation text to visualisations Available on web and mobile 2 Hard to read [1] London Air: http://www.londonair.org.uk/london/asp/PublicEpisodes.asp [2] Defra: http://uk-air.defra.gov.uk/latest/ [3] London Air App: http://itunes.apple.com/gb/app/london-air/id358970517?mt=8
  • 8. 2.Numeric Data [1] [2] [3] [1] The London Marylebone Road AQI: http://aqicn.org/city/united-kingdom/london-marylebone-road/m/hk/&aboutaqi [2] China Air Quality Index App: https://itunes.apple.com/us/app/china-air-quality-index/id477700080?mt=8 [3] China Air App: https://itunes.apple.com/us/app/chinas-air/id777458271?mt=8
  • 9. 2.Numeric Data [1] [2] [3] 1 Difficult to show spatial & temporal variations 1 Good to show current air quality [1] The London Marylebone Road AQI: http://aqicn.org/city/united-kingdom/london-marylebone-road/m/hk/&aboutaqi [2] China Air Quality Index App: https://itunes.apple.com/us/app/china-air-quality-index/id477700080?mt=8 [3] China Air App: https://itunes.apple.com/us/app/chinas-air/id777458271?mt=8 Combined with other kinds of visualisations 2 Use numeric data, colour blocks, health advice 2 People don’t understand “moderate” air quality Use photographs to add an emotional connection to data Use healthy, unhealthy, understandable words
  • 10. 3.Line Graph [1] 1 InAir [1]! ! • People were interested to see spikes in visualisations. ! • Give users data and let them decide what to do Difficult to show spatial variations 1 Good to show temporal variations [1] Kim, S., Paulos, E. & Mankoff, J. (2013). inAir: A Longitudinal study of Indoor Air Quality Measurements and Visualizations. CHI 2013, April 27 – May 2, 2013, Paris, France.
  • 11. Public Usage People are not aware of the existing air quality information Personal! Usage Channels: web, newspaper, mobile app ! Format: text, map, numeric data and line graph. 2 Designed based on the low spatial and/or temporal resolution data set in cities 1 EXISTING VISUALISATIONS
  • 12. Public Usage Personal! Usage Channels: web, newspaper, mobile app ! Format: text, map, numeric data and line graph. People are not aware of the existing air quality information 2 Designed based on the low spatial and/or temporal resolution data set in cities 1 Not well applied to ! a mobile context 3 EXISTING VISUALISATIONS
  • 13. No personalised data 1 Didn’t show temporal variations 2 Didn’t address on the NO2 pollution 3 Depressing to check data 4 5 Not playful EXISTING MOBILE VISUALISATIONS [1] [2] [3] [1] AQ Scotland App: https://itunes.apple.com/en/app/air-quality-in-scotland/id838197830 [2] City Air App: https://itunes.apple.com/gb/app/city-air/id706049131?mt=8 [3] Birkett Index App: https://itunes.apple.com/gb/app/birkett-index/id646281816?mt=8
  • 14. AIR QUALITY INDEX • AQI is widely used in air quality visualisations • Inconsistent between countries • No standard threshold to define the healthy and unhealthy levels. No thresholds of health effects can be identified (Laden et al, 2006)! Because people’s susceptibility to pollution levels varies, and health effects on different people also varies. The UK AQI: SOURCE: Laden, F., Schwartz, J., Speizer, F. E., & Dockery, D. W. (2006). Reduction in fine particulate air pollution and mortality - Extended follow-up of the Harvard six cities study. American Journal of Respiratory and Critical Care Medicine, 173(6), 667-672. doi: Doi 10.1164/Rccm.200503-443oc
  • 15. Policy-defined NO2 limits for cities. ! No standard for citizen exposure is available. EU Annual Mean Limit (Healthy for general population) 40 μg/m3 EU Hourly Mean Limit UK AQI Low Band 200 μg/m3 (Healthy for at-risk group) UK AQI Moderate Band 200 μg/m3 400 μg/m3 (Not to be exceeded 18 times per year) COMEAP, Committee on the Medical Effects of Air Pollutants (2011). Review of the UK Air Quality Index, A report by the Committee on the Medical Effects of Air Pollutants. Retrieved from http://www.hpa.org.uk/webc/HPAwebFile/HPAweb_C/1317137023144 Air Quality Standards from European Commission. Retrieved from http://ec.europa.eu/ environment/air/quality/standards.htm
  • 17. People’s susceptibility to pollution levels varies! ! Volumes of pollutants vary temporally. 1 2
  • 18. ! It is difficult to decide on a healthy/unhealthy volume.
  • 19. ! It is difficult to decide the measured length of the averaging period.
  • 20. VISUALISATIONS PLAYFUL WearAir Air quality information was visualised not only to the wearer but also to the people around, which shows the importance of making air quality visualisation playful and engaging in order to raise the users’ initiative to share the visualisation with others. SOURCE: Kim, S., Paulos, E. & Gross, M. (2010). WearAir: Expressive T-shirts for Air Quality Sensing, TEI 2010, January 25–27, 2010, Cambridge, Massachusetts, USA
  • 21. Balloons Participants were encouraged to take balloons with them to different parts of the city to explore air quality by themselves. They were interested and surprised to see changes of air quality and came up with ideas on why the air quality was changing regarding to their behaviours. VISUALISATIONS PLAYFUL SOURCE: Kuznetsov, S., Davis, G., Paulos, E., Gross, M. and Cheung, J. (2011). Red Balloon, Green Balloon, Sensors in the Sky. UbiComp 11, September 17–21, 2011, Beijing, China.
  • 22. Floating Beijing SOURCE:Guler, D. & Wang, X. (2012). FLOAT Beijing. Retrieved from http://f-l-o-a-t.com/ Because people were finding it hard to see the stars in the Beijing night sky, they visualised things that people should be able to see but could not see due to pollution, and provided a strong visual contrast. VISUALISATIONS PLAYFUL
  • 23. [1] [2] [3] • Simple visualisations ! ! • Visual contrast! ! • Playful and engaging! ! • Show changes of AQ data VISUALISATIONS PLAYFUL [1] Kim, S., Paulos, E. & Gross, M. (2010). WearAir: Expressive T-shirts for Air Quality Sensing, TEI 2010, January 25–27, 2010, Cambridge, Massachusetts, USA [2] Kuznetsov, S., Davis, G., Paulos, E., Gross, M. and Cheung, J. (2011). Red Balloon, Green Balloon, Sensors in the Sky. UbiComp 11, September 17–21, 2011, Beijing, China. [3] Guler, D. & Wang, X. (2012). FLOAT Beijing. Retrieved from http://f-l-o-a-t.com/
  • 24. [1] [2] [3] • Data from mobile sensors! ! • Not quality data VISUALISATIONS PLAYFUL [1] Kim, S., Paulos, E. & Gross, M. (2010). WearAir: Expressive T-shirts for Air Quality Sensing, TEI 2010, January 25–27, 2010, Cambridge, Massachusetts, USA [2] Kuznetsov, S., Davis, G., Paulos, E., Gross, M. and Cheung, J. (2011). Red Balloon, Green Balloon, Sensors in the Sky. UbiComp 11, September 17–21, 2011, Beijing, China. [3] Guler, D. & Wang, X. (2012). FLOAT Beijing. Retrieved from http://f-l-o-a-t.com/
  • 25. SENSORS FIXED LOCATION LAQN Data Set Regulatory grade sensors Sensing London Project Sensor Cost High cost Lower cost Data Quality High quality & calibrated High quality & calibrated Spatial Resolution of Data Low spatial resolution High spatial resolution Temporal Resolution of Data Low temporal resolution High temporal resolution LAQN: http://www.londonair.org.uk/LondonAir/Default.aspx Sensing London project: https://futurecities.catapult.org.uk/project-full-view/-/asset_publisher/oDS9tiXrD0wi/content/project-sensing-london/
  • 26. Reach a wider audience PEOPLE Show changes of ! air quality data 4 Give people air quality data, and let them decide what to do VISUALISATIONS PLAYFUL Visual objects, or things that people could not see because of pollution Simple & understandable ways Interactive lights, and by changing colours (red, yellow and green), volumes or by blinking 1 2 Playful and engaging ways Not only to the user but to people around 3 Visual contrast
  • 27. Visual objects, or things that people could not see because of pollution Simple & understandable ways Interactive lights, and by changing colours (red, yellow and green), volumes or by blinking 1 2 Playful and engaging ways not only to the user but to people around 3 Visual contrast Show changes of air quality data 4 Give people air quality data, and let them decide what to do Fit into people’s daily lives 5 Use habitual behaviours VISUALISATIONS REQUIREMENTS
  • 28. PERCEIVED INDICATORS Air Quality! Understandable PERSONAL! EXPERIENCE NOTICE POLLUTION Visual cues, odours, residue, health effects Trigger interests / read more articles / come up with solutions to avoid or reduce pollution USER INTERVIEWS FINDINGS
  • 29. Participants understood the air quality was bad via visual cues. Photographs
  • 30. [2] [3] [1] [1] The Atlantic (2012). A Stunning Visualization of China's Air Pollution. Retrieved from: http:// www.theatlantic.com/ international/archive/2012/07/ a-stunning-visualization-of-chinas- air-pollution/259455/ [2] China Air Daily: http:// www.chinaairdaily.com/ [3] Guler, D. (2012). FloatPM. Interactive Art and Computational Design. Retrieved from http:// golancourses.net/2012spring/ 02/09/deren-guler_ project1_float-pm/ VISUALISATION PHOTOGRAPHS
  • 31. [2] [3] [1] [1] The Atlantic (2012). A Stunning Visualization of China's Air Pollution. Retrieved from: http://www.theatlantic.com/international/archive/2012/07/ a-stunning-visualization-of-chinas-air-pollution/259455/ [2] China Air Daily: http://www.chinaairdaily.com/ [3] Guler, D. (2012). FloatPM. Interactive Art and Computational Design. Retrieved from http://golancourses.net/2012spring/02/09/deren-guler_ project1_float-pm/ 1 Used original photos Addressed on the PM pollution in Beijing 2 3 No data over photographs 1 Comparison is a key to help people understand 1) To understand if the current data 2) To understand the trend of air quality But NO2 is less visible… VISUALISATION PHOTOGRAPHS
  • 32. “When I saw the number I can feel how good or bad the air quality level will be for today, because I’m used to it.” (P7) — A Beijing resident, using the app for two years Interpret the level of severity of air pollution to the experience in real life from reading the AQI number. USER INTERVIEWS FINDINGS China Air Quality Index App: https://itunes.apple.com/us/app/china-air-quality-index/id477700080?mt=8
  • 33. Make the invisible visible Display data + DESIGN CONCEPT
  • 34. Participants want to see… Temporal! Variations Changes & trends! in past 10 years? Spatial! Variations ! at my home? ! at my work place?! in Beijing? How is the air ! quality at my ! current location? Health effects? The safe/dangerous level? Actions can take? Causes?
  • 36. CAMERA! VIEW Take a photo APPLY! FILTERS SHARE! THE PHOTO MY PHOTO! STREAM INSTANO2 MODEL
  • 37. HEALTHY AND UNHEALTHY THRESHOLD The EU annual mean value of 40 μg/m3 ! INFORMATION: Numeric data, colour block, a healthy/unhealthy label the user’s current location and time Only NO2 data INSTANO2 VISUALISATION
  • 38. Origin! photo INSTANO2 VISUALISATION Make the invisible visible via photographs Adjust the appearance of the photograph based on the air quality data
  • 39. INSTANO2 CAMERA VIEW Take a photo to explore air quality The current NO2 level at the current location
  • 40. INSTANO2 LOCATION FILTER Compare the current air quality data with OTHER PLACES: Places that people are familiar with, such as places in London, major cities in the UK, capital cities in Europe and over the world.
  • 41. INSTANO2 TIME FILTER Compare the current air quality data with HISTORICAL DATES: Yesterday, last week, last month, the same day in past months of this year, and the same date in past years.
  • 42. INSTANO2 SHARING 1 Talk and share air quality data 2 3 Social media Photos collection website My Places
  • 43. Real time air quality data at places,! home, work places, places people passing by everyday. Motivate people to check air quality Like checking weather at places INSTANO2 MY PLACES
  • 44. Experimental area, LAQN data set, low spatial and temporal resolution Take six pictures using the app Participants were encouraged to go anywhere they liked to and the task was to take six pictures using the app. USER EXPERIMENTS DESIGN
  • 45. USER EXPERIMENTS FINDINGS “ Places participants went to Main streets! Took photos with lots of traffic, especially with buses, expecting to see poor air quality. Places they went to everyday! Gyms where they swam, streets they frequently walked on, shops where they grabbed sandwiches, cafes where they had lunch and parks where they relaxed. To understand the air quality they were exposed to every day. Places they’ve never been to! Cafes and hospitals nearby, to check whether they should go there in the future, to explore good spots to go to. 1 2 3
  • 46. USER EXPERIMENTS FINDINGS “ Places where they would like to use the app 1 2 3 Home and work/study places! Check air quality frequently ! Places with lots of traffic and people! Oxford Street and Regent Street. ! Tourist places! River Thames and Kew Garden. These were places where they would usually take pictures and by using the app, they could also check if the air quality was good or bad there.
  • 47. USER EXPERIMENTS FINDINGS “People like taking photos. It’s easy to do.” (E3) “I like the camera interface. Users don’t have to learn a new interface.” (E4) The app was easy to use. Participants were using the app just like using other camera apps.
  • 48. USER EXPERIMENTS FINDINGS “I think colour is enough to represent the current air quality. But photos made it more real, and give you a sense of how good or bad it might affect you. When I see the current picture applied with a poor air quality filter, I feel dirty and a little uncomfortable, and I want to move away from that place soon.” (E5) Photographs visualised the effects of air pollution that could affect him.
  • 51. Location filters Understand the spatial variations of air quality • Search function • Photo collage USER EXPERIMENTS FINDINGS
  • 52. Time filters “I noticed the air quality is far better in the winter.” (E7) • Understand temporal variations of air quality • Reflect in humorous ways USER EXPERIMENTS FINDINGS
  • 53. Pre-study interviews During Experiments Depressing to check data Participants reflected in humorous ways! Have no idea how air quality data looks like even when they were facing poor air quality. The app sparked curiosity, triggered exploration and raised awareness Participants thought and talked about what air quality meant, where the pollution came from, what might change the air quality and why it was changing. They also asked for more explanations about the data, expressed their own initiative to interpret and explore the data further to learn more. USER EXPERIMENTS FINDINGS
  • 54. Data is lacking regarding physical locations and not dynamic. Experimental area, LAQN data set, low spatial and temporal resolution Difficult to show changes DESIGN CHALLENGES
  • 55. DESIGN CHALLENGES SENSOR 1 SENSOR 2 27 110 CHANGES 1:! Participants can see changes of data when walk in between sensors! Data source changes from one sensor to the other.
  • 56. CHANGES 2: On camera view, show a more dynamic data by adding a random variation on data. Data from the nearest sensor Data varies by a range of 10 by every 4 seconds X [X, X+10] 100 [100, 110] 55 [55, 65] DESIGN CHALLENGES
  • 57. USER EXPERIMENTS FINDINGS One participant lost interest using the app Need a higher spatially dense data set DESIGN OPPORTUNITY Request Sensors! The app says data not available, or that the nearest sensor is 5 miles away, the user could press a button and request city authorities to deploy sensors in that area. DESIGN OPPORTUNITY Use computer vision! ! Recognise the content in the photo, and adjust NO2 levels based on numbers of cars, buses, etc.. Because he noticed the low spatially dense data set, changes in data didn’t match with his expectations
  • 58. Public Usage Personal Usage More customisation PUSH • Search function for locations, make photo collages using the app. • Highlight key event dates in the time filters, such as the introduction of the congestion charge in London. More spatially dense data set INSTANO2 FUTURE WORKS Air Quality Data from fixed location sensors Users • More rational spikes can be seen from the data. A longer experiment • to learn about their motivations to check air quality data and their behaviour changes. 1 2 3
  • 59. Public Usage Personal Usage GENERATE INSTANO2 FUTURE WORKS 1 User usage data (time/location data) Air Quality Data from fixed location sensors PUSH Users 2 Photo-based data USE Other stakeholders, such as city officials
  • 60. Public Usage Personal Usage 1 User usage data (time/location data) Data to city officials Where Where citizens are interested to see air quality data When When people are using the app Who Who these users are Locations where air quality sensors are needed. Timing when air quality data needs to be pushed to the users.. Target Group to whom air quality is important. INSTANO2 FUTURE WORKS
  • 61. Public Usage Personal Usage INSTANO2 FUTURE WORKS 2 Photo-based data add value to numerical data from the sensors Photos Photos collection • #searchable • sharable • contextual • Visualisation of London’s air quality • Crowdsourced insights from citizens • To learn citizens’ conversations about air quality
  • 63. Visualisation of London’s air quality! Crowdsourced insights from citizens To learn citizens’ conversations about air quality
  • 64. Thanks! Q&A? Ye Lin Supervisors: Lorna Wall, Han Pham | 09/2014