IS428 AY2019-20T1 Assign Kok Jim Meng Tasks Questions and Answers
IS428 VAST 2019 Mini Challenge 2
|
|
|
|
|
|
Tasks Questions and Answers
Task 1
Dashboard
The dashboard provides a total of 3 visuals:
1. Map of the Mobile Sensors and Static Sensors throughout the period from 6 April 2020 to 10 April 2020
Map that provides the location of the sensors of their whereabouts since 6 April 2020 to 10 April 2020. Serves as a filter to filter the other charts. Hover to see the sensor id, its radiation value, the location coordinates, and the timestamp in minutes.
2. Radiation Measurement (Maximum) for Mobile Sensors in Minute of the Timestamp
A time-series chart based on minutes by using a sample size of maximum radiation measurement value from 0 to 5000 of the mobile sensors. Marked by colour intensity where light blue is low and dark blue is high. Hover to see the sensor id and its radiation value.
3. Radiation Measurement (Maximum) for Static Sensors in Minute of the Timestamp
A time-series chart based on minutes by using a sample size of maximum radiation measurement value from 0 to 5000 of the static sensors. Hover to see the sensor id and its radiation value.
Task 1 Questions & Answers
Visualize radiation measurements over time from both static and mobile sensors to identify areas where radiation over background is detected. Characterize changes over time. Limit your response to 6 images and 500 words.
Description |
Images |
|
When the radiation measurement of a sensor has reached a minimum value of 100 CPM, it is at the dangerous radiation level. According to the International Nuclear and Radiological Event Scale, there are 8 classes of radiation classification as follows:
Hence, we will choose “6. Accident with wider consequences”, “7. Serious accident”, and “8. Major accident” to find out where do these radiation classifications are detected as these classifications are range from 100 CPM onwards. In the map, the blue dots indicate the mobile sensors while the orange dots indicate the static sensors. Based on the timeline of April 6 2020 12.37am to April 10 2020 11.59pm, most of the high radiation occur in the north and central part of St Himark such as:
This is most likely that the radioactive contamination has spread from the nuclear plant from Safe Town to its nearby regions. |
 |
|
While selecting M-9 during that timestamp as a filter in the dashboard, sensor M-9 is actually near the nuclear plant in Safe Town. We assume that since that timestamp strikes, the radiation contamination has started to take place as according to the time-series chart. |
 |
|
With the same indication of the radiation classification as mobile sensors’, for the static sensors, only S-15 has the high radiation measurement as the sensor is fixedly placed near the nuclear plant in Safe Town after selecting S-15 in the time-series chart of static sensors as the filter in the dashboard. |
 |
|
When indicating all the radiation classification for the static sensors, as compared to the rest of the static sensors, only S-15 has no value from 8 April 2020 10.06pm and the value has resumed from 10 April 2020 8.45pm onwards. Probably during that period, S-15 is faulty as it is the only static sensor that is placed very near to the nuclear power plant and most likely both earthquake and the plant leakage were occurred during that period. |
 |
Task 2
Dashboard 1
This dashboard answers the first 2 parts of the task and it provides a total of 3 visuals:
1. Contamination Path
Map that provides the location of the sensors of their whereabouts since 6 April 2020 to 10 April 2020 in hours. Hover to see the sensor id and its sensor type, its radiation value, the location coordinates, and the timestamp.
2. Mobile Anomaly and Uncertainty
A time-series line chart using a sample size of maximum radiation measurement value from 0 to 5000 showing the mobile sensors’ id and its maximum radiation measurement. This chart helps to check where is the uncertainty and the anomaly. Hover to see its value and the timestamp. It is used as a filter for the contamination path chart.
3. Static Anomaly and Uncertainty
Similar to Mobile Anomaly and Uncertainty, it is also a time-series line chart using a sample size of maximum radiation measurement value from 0 to 5000 showing the mobile sensors’ id and its maximum radiation measurement. This chart helps to check where is the uncertainty and the anomaly. Hover to see its value and the timestamp. It is used as a filter for the contamination path chart.
Dashboard 2
This dashboard answers the last part of the task and it provides a total of 4 visuals which all of them serve as filters:
1. Map
A map of the sensors’ whereabouts. Hover to see sensor-id and its location coordinates.
2. Time Series (Minutes) of Radiation Measurements by both Mobile and Static Sensors
Time-series line chart showing the effects of happenings in St Himark over the period. Use a sample size of maximum radiation measurement from 0 to 5000. Hover to see the radiation measurement and the timestamp.
3. Mobile Anomaly and Uncertainty
A time-series line chart using a sample size of maximum radiation measurement value from 0 to 5000 showing the mobile sensors’ id and its maximum radiation measurement. This chart helps to check where is the uncertainty and the anomaly. Hover to see its value and the timestamp. It is used as a filter for the contamination path chart.
4. Static Anomaly and Uncertainty
Similar to Mobile Anomaly and Uncertainty, it is also a time-series line chart using a sample size of maximum radiation measurement value from 0 to 5000 showing the mobile sensors’ id and its maximum radiation measurement. This chart helps to check where is the uncertainty and the anomaly. Hover to see its value and the timestamp. It is used as a filter for the contamination path chart.
Task 2 Questions & Answers
Use visual analytics to represent and analyze uncertainty in the measurement of radiation across the city. Limit your responses to 12 images and 1000 words.
1. Compare uncertainty of the static sensors to the mobile sensors. What anomalies can you see? Are there sensors that are too uncertain to trust?
2. Which regions of the city have greater uncertainty of radiation measurement? Use visual analytics to explain your rationale.
Description |
Images |
|
As discussed in Task 1 while indicating the radiation classification with at least 100 CPM, S-15 has no value from 8 April 2020 10.06pm and the value has resumed from 10 April 2020 8.45pm onwards. This is due to it is fixedly placed very near to the nuclear plant. In this case when all radiation classification takes into account, S-15’s value has stopped at 20 CPM at 10.06pm on 8 April 2020 until when it reaches at 8.54pm on 10 April 2020 the value has risen to 24 CPM. As compared to the rest of static sensors, there are spikes during that period. However, at the point in time of 10.06pm on 8 April 2020, the radiation value of S-1, S-6, S-14 lay low until the morning and afternoon of 9 April 2020. |
 |
|
While selecting the 3 static sensors in the Static Anomaly and Uncertainty chart as the filter, S-1 is placed at Palace Hills, S-6 is placed at Southwest, and S-14 is placed at Cheddarford. We assume that earthquake has stroked in these areas during that period. This will be discussed in later parts of this Task 2. |
 |
|
As for the anomaly of the static sensors, before the earthquake and the leakage period (as labelled in the blue square which also discussed about the uncertainty previously), S-1, S-11, and S-12 have anomalies as the spikes happened at a different period as unlike to the other static sensors where the spikes happened in a fixed period. |
 |
|
In terms of uncertainty for mobile sensors, as there are 50 mobile sensors, I have picked 10 sample mobile sensors (M-1, M-5, M-8, M-12, M-14, M-16, M-17, M-35, M-41, M-50) which seem uncertain. Sensors such as M-1 and M-35, they have a few of spikes before the morning of 8 April 2020 and thereafter their radiation values lay low. As for M-8, M-17, and M-41, their values lay low until the morning of 8 April 2020 when they have spikes of radiation measurements. Lastly, for M-5, M-12, M-14, M-16, and M-50, there are values lay low after the morning of 7 April 2020 and continue to have spikes in the morning of 8 April 2020. From above we can see that the sensors of M-5, M-8, M-12, M-14, M-16, M-17, M-41, and M-50 have spikes after the morning of 8 April 2020. Hence, we will explore more on the uncertainty of M-1, M-35, and a few more sensors that follow the same trend as them. |
 |
|
We have picked M-1, M-6, M-26, and M-35 from the Mobile Anomaly and Uncertainty chart as a filter to show the contamination path in the map of the 4 mobile sensors. Based on the map, the paths are at these regions:
Furthermore, as explained earlier in the uncertainty for static sensors, Palace Hills, Southwest, and Cheddarford were also involved. Hence, we assume that earthquake has stroked the above regions and sensors in Palace Hills, Southwest, and Cheddarford are the most uncertain to trust. |
 |
|
By using the same 10 sample size of the mobile sensors to explain the anomaly, before the earthquake and the leakage period (as labelled in the blue square which also discussed about the uncertainty previously), M-1, M-5, M-12, M-14, M-16, M-35, and M-50 have anomalies. |
 |
|
Due to these sensors have anomaly, we picked them as filter to see where these mobile sensors are placed in the region and from what we have observed, most of them are located in the northwest of St Himark which are very near to the Safe Town region where the nuclear plant is located. We can assume that once the nuclear plant leaks, those regions in the northwest of St Himark will be affected first. |
 |
3. What effects do you see in the sensor readings after the earthquake and other major events? What effect do these events have on uncertainty?
Description |
Images |
|
On April 8 2020, between 1.15pm to 4.40pm, there is foreshock (labeled in red rectangle). Thereafter, from 4.40pm to 10.15pm, there is earthquake (labeled in brown rectangle). Furthermore, the data given in VAST 2019 Mini Challenge 1 has also shown that on 8 April 2020, there are high shake intensity. This means that earthquake has occurred during this period. After the earthquake, there is leakage period (labeled in purple rectangle) and the nuclear plant has started to leak (left white box) at 2.40pm on April 9 2020 then it rises steadily at 6.39pm and drop at 7.56am on April 10 2020 and the leakage period ends at 1.10pm. |
 |
|
Due to the size of the dashboard, the Mobile Anomaly and Uncertainty chart is not in Entire View but in Fit Width instead so that can see the charts easily and can be scrolled. While selecting the period between the earthquake and leakage period in the time series chart as filter, mobile sensors M-1, M-11, M-23, M-26, and M-47 do not really have spikes. Moreover, majority of the mobile sensors value went to null at a certain period of time but except for M-18 which has null value and at the certain period it has values with spikes. As for the static sensors, it is quite reliable as most of the readings are standardized except for S-15 where its value is cut at 10.06pm on April 8 2020. The radiation values start to rise after the earthquake where the sensors are located in the northwest and southeast of St Himark which are near to the Safe Town region where the nuclear plant is located. This means the nuclear plant is going to start leaking. |
 |
|
While selecting the period after the leakage period in the time series chart as filter, majority of the mobile sensors are either having null value at the start and at the certain period it has values with spikes or it has values until a certain period the values went to null. The values of M-1, M-26, M-35, M-47 do not really have spikes. As for the static sensors, it is quite reliable as most of the readings are standardized except for S-15 where its value went null until 8.45pm on April 10 2020 when the values arises. After the leakage period, most of the sensors are located in the northwest and southeast of St Himark which are near to the Safe Town region where the nuclear plant is located. This means the radiation contamination has already happened. |
 |
Task 3
Dashboard 1
This dashboard answers the first part of the task and it provides a total of 1 visual:
1. Contamination Path
Map that provides the location of the sensors of their whereabouts since 6 April 2020 to 10 April 2020 in hours. Hover to see the sensor id and its sensor type, its radiation value, the location coordinates, and the timestamp.
Dashboard 2
This dashboard answers the first part of the task and it provides a total of 1 visual:
1. Contaminated Cars during Leakage Period
Map that provides the location of the sensors of their whereabouts from 6.39pm April 9 2020 to 7.09am April 10 2020 which is the leakage period which also serves as a filter. The colours represent the different sensors. Use Radiation Classification filter as a gauge to see how danger the radiation is during the selected timing of the leakage period. Hover to see the sensor id and its sensor type, its radiation value, the location coordinates, and the timestamp.
Dashboard 3
This dashboard answers the last part of the task and it provides a total of 2 visuals:
1. Enough Mobile Sensors?
Map that provides the location path of the mobile sensors during each day which is classified into different colours. Use timestamp filter to check whether there are enough sensors for one of the days and use Radiation Classification filter to see that particular radiation level has enough sensors. Hover to see the sensor id and its sensor type, its radiation value, and the location coordinates.
2. Enough Static Sensors?
Map that provides the location of the static sensors. Use timestamp filter to check whether there are enough sensors for one of the days and use Radiation Classification filter to see that particular radiation level has enough sensors. Hover to see the sensor id and its sensor type, its radiation value, and the location coordinates.
Task 3 Questions & Answers
Given the uncertainty you observed in question 2, are the radiation measurements reliable enough to locate areas of concern? Limit your responses to 12 images and 1000 words.
1. Highlight potential locations of contamination, including the locations of contaminated cars. Should St. Himark officials be worried about contaminated cars moving around the city?
Description |
Images |
|
By selecting all of the sensors, the map shows all the paths of the sensors throughout the period of time from 6 April 2020 to 10 April 2020. Based on the map, because of high radiation measurement, potential locations of contamination include:
The officials should be worried as regions that are far away from Safe Town, where the nuclear plant is located, such as Broadview, Scenic Vista, and Terrapin Springs are identified as potential locations of contamination as well. |
 |
2. Estimate how many cars may have been contaminated when coolant leaked from the Always Safe plant. Use visual analysis of radiation measurements to determine if any have left the area.
Description |
Images |
|
The leakage period is started from 6.39pm 9 April 2020 to 7.09am 10 April 2020. From the map, we assumed that at first, once there is a leak, Safe Town will be the first region to bring contamination to nearby regions as the nuclear plant is in Safe Town. However, during the period of the leakage, both M-13 and M-32 are very near to the nuclear plant in Safe Town at 7.32pm and 8.25pm respectively on 9 April 2020. M-32 only moves from Safe Town and stops in Pepper Mill at 5.33am 10 April 2020 without bringing any contamination to other cars while M-13 didn’t leave Safe Town from 7.32pm on 9 April 2020 to 7.09am on 10 April 2020. |
  |
|
Actually, most of the cars have already been contaminated since the morning of 10 April 2020. |
 |
|
For instance, M-43 has been contaminated in West Parton at the start of the leakage period and brought contamination to Easton. |
 |
|
Same goes to M-39 which has brought contamination from Southton to West Parton to Easton to Safe Town. And as for M-36, it brought contamination from Easton to Old Town. |
 |
|
As such, potential contamination location that arises from will be in Easton. Moreover, most of the contamination path are started from Easton in the morning of 10 April 2020 to other regions till the end of the leakage period. |
 |
3. Indicated where you would deploy more sensors to improve radiation monitoring in the city. Would you recommend more static sensors or more mobile sensors or both? Use your visualization of radiation measurement uncertainty to justify your recommendation.
Description |
Images |
|
We choose “6. Accident with wider consequences”, “7. Serious accident”, and “8. Major accident” as filters because these classifications are range from 100 CPM onwards. As such, using these radiation classifications allow us to know where should be the potential locations to put more sensors at. As time goes by from 6 April 2020 to 10 April 2020 while both earthquake and leakage have happened, there are enough mobile sensors to detect the radiation in the northwest of St Himark where the regions are shared their boundary with Safe Town. However, in the southeast of St Himark (Broadview, Chapparal, Scenic Vista, Terrapin Springs), the number of mobile sensors start to drop when time goes by. Hence, there is a need to put more mobile sensors in the southeast of St Himark. As for static sensors, no static sensors were placed in the central of St Himark (Weston, Easton, Southton, West Parton, East Parton, Oak Willow). We believe that static sensors have to be placed in these regions to understand if high radiation occurred in this area as well. |
 6 April 2020
 7 April 2020
 8 April 2020
 9 April 2020
 10 April 2020
|
Task 4
Dashboard
The dashboard provides a total of 3 visuals:
1. Map of the last 4 hours of 10 April 2020
Map that provides the location of the sensors of their whereabouts during the last 4 hours of 10 April 2020. Serves as a filter to filter the other charts. Hover to see the sensor id, sensor classification, and the location coordinates.
2. Maximum Radiation Measurement for Mobile Sensors in Minute of the Timestamp on the last 4 hours of 10 April 2020
A time-series chart of the last 4 hours of 10 April 2020 based on minutes by using a sample size of maximum radiation measurement value from 0 to 5000 of the mobile sensors. Marked by colour intensity where light blue is low and dark blue is high. Hover to see the sensor id, timestamp, and its radiation value.
3. Maximum Radiation Measurement for Mobile Sensors in Minute of the Timestamp on the last 4 hours of 10 April 2020
A time-series chart of the last 4 hours of 10 April 2020 based on minutes by using a sample size of maximum radiation measurement value from 0 to 5000 of the static sensors. Hover to see the sensor id, timestamp, and its radiation value.
Task 4 Questions & Answers
Summarize the state of radiation measurements at the end of the available period. Use your novel visualizations and analysis approaches to suggest a course of action for the city. Use visual analytics to compare the static sensor network to the mobile sensor network. What are the strengths and weaknesses of each approach? How do they support each other? Limit your response to 6 images and 800 words.
Description |
Images |
|
According to the map, the strength of mobile sensors is that they can move constantly in a region which allow us to monitor the radiation dynamically. For example, in Old Town, we can see that different mobile sensors are moving in a certain direction during the last 4 hours of the last day. However, the weakness of mobile sensors is that they will bring contamination to regions that have lower risks of getting contaminated by the nuclear leakage. The strength of static sensors is that they are fixed placed in their respective locations which means their radiation measurement are less likely to be affected by external factors. However, their weakness is that the coverage of the radiation measurement is limited as they are only placed in a few regions such as Palace Hills, Southwest, Downtown, Old Town, Safe Town, Cheddarford, and Broadview. Based on the map, we can suggest to place static sensors in regions with less mobile sensors such as Oak Willow, Chapparal, Pepper Mill, and Terrapin Springs. This allows us to further analyse those regions’ radiation measurement. Furthermore we can analyse more on the mobile sensors based on their most frequently visited regions as most of the them are in the northeast of St Himark. |
 |
|
As mentioned earlier that static sensors’ radiation measurement are less likely to be affected by external factors, the chart has evidently shown due to the certain radiation measurement. However, there are a total of 9 static sensors but during the last 4 hours on the last day, there are only 8 static sensors as S-15 went missing which most likely faulty after earthquake and leakage period as it is initially placed very close to the nuclear plant in Safe Town. |
 |
|
In addition to the weakness for mobile sensors, the readings of mobile sensors are probably affected by external factors such as the condition of roads, or the materials of the cars. The chart shows that some mobile sensors have high values at the start and at the end of the last 4 hours of the last day but with empty values in between the period of 4 hours. Probably those sensors are already contaminated or have become faulty due to high radiation measurements. Furthermore, these sensors are the outskirts of southeast of St Himark – Broadview, and Scenic Vista Based on the map, we can suggest to place more mobile sensors or static sensors the above 2 regions to analyse why these regions have high and yet null values during the last 4 hours of the last day. |
  |
Task 5
Task 5 Questions & Answers
The data for this challenge can be analyzed either as a static collection or as a dynamic stream of data, as it would occur in a real emergency. Describe how you analyzed the data – as a static collection or a stream. How do you think this choice affected your analysis? Limit your response to 200 words and 3 images.
When using stream data, stream processing is needed and the goal of using stream data is to analyse data as close to the event location as possible and decide on the critical timestamp when applying to this challenge. Anyhow, analysing using stream data helps to identify and gain as much insights as possible and examine patterns, especially the latest data points, when events occur so that we can respond quickly to alleviate the problems.
Stream processing fits well with time-series data and detecting patterns over time. In application to this challenge, we can use time-series charts such as calendar heatmap showing the relationship between radiation values of sensors and time, and cartograms or event heatmap of St Himark map to show regions with high radiation contamination. These real-time charts will be placed in a visualisation.
As for static data, they might have null values, outliers, and messy data. Hence, there is a need to clean and processing the data before putting them into analysis. Static data allows us to analyse the overview of the situations during the period of time which is given to us. What we have done so far for this challenge is analysing using static data.
