Background: Overwhelmed by the scale of the disaster, city of Himark officials are looking for help to build a more complete understanding of the situation on the ground. They have access to multiple data sources, but each on its own is an incomplete view into the situation. Fuse the data in the mini challenges to give officials a fuller situational awareness than is possible with any individual data sources. Officials are especially interested in understanding how the uncertainty in each source interacts with the others. Can sources reinforce or refute findings in other sources? Does combining sources bring the situation into clearer view or just increase confusion? The citizens of St. Himark are trying to assist their government by providing more information than official channels would be able to on their own, but you must help fuse those sources into a form that allows officials to do their job as efficiently and with the greatest impact possible.

Task and Questions

Q1

Generate a master timeline of events and trends during the emergency response. Indicate where it is uncertain and which data underlies that uncertainty. Limit your response to 1000 words and 12 images.

Answer:

A master timeline:

This is a brief timeline, each marked point stands for a specific event, from MC 1, 2 or 3, y-Axis is the importance of events.

There are 3 points on y = 0, which means the start times of each earthquake.

We list all the events in the table below, ordered by their start time.

Q2

Identify and explain cases where data from multiple mini-challenges help to resolve uncertainty, and identify cases where data from multiple mini-challenges introduces more uncertainty. Present up to 10 examples. If you find more examples, prioritize those examples that you deem most relevant to emergency response. Limit your response to 1000 words and 12 images.

Answer:

Cases where data from multiple sources help to resolve uncertainty:

1. MC3 reports an earthquake 14:33 Apr. 6th and damage records from MC1 approve it.

2. Sudden jumps were detected by static sensors S9, S12 at 4/9 14:24, and other static sensors showed little higher radiation level right after that in MC2’s data. Other data from MC1 report damage increment at 13:00 while data from MC3 show close of bridge at the same time. All these data indicate earthquake happened at the point and give reasons to these anomalies.
3. MC3’ data shows that the HSS group claimed to have a meeting at 19:00 Apr. 7th. Just the next morning, the striped patterns of mobile sensors disappeared in MC2’s data. One possible reason is that they discussed at the meeting and then changing the sensors’ setting.
4. MC1 reports the third earthquake at 15:00 Apr.9th, and 15:06, someone sent messages to ask if there was an earthquake since the wall moved.
5. MC1 reports that Old Town loses connection and at the same time MC2’s data indicates that the static sensors’s reading went higher.
6. MC1 claims that Old Town suffer the most during the earthquake and MC3’s messages approve it with plenty of homeless people and short of food and water. People were suggested to leave the area.
7. Someone said in MC3’s data that nuclear reactor was witnessed at Scenic Vista and MC2 shows that mobile sensor 20 recorded abnormal radiation level before got lost.

sensor 20’s pattern from MC2

8. Reports from MC1 on damage of roads and bridges came should be followed by temporary closed of these bridges and highways.

highways and bridges’ information’s frequency from MC3

reports about roads in District 12, from MC 1

Cases where data from multiple sources introduces more uncertainty:

1. MC3 reports earthquake at 08:36 Apl. 8th, while MC1’s damage records starts from 08:00
2. The time MC2’s sensors’ striped patterns disappear do not correspond with the time earthquake happened.

Q3

Are there instances where a pattern emerges in one set of data before it presents itself in another? Could one data stream be used to predict events in the others? Provide examples you identify. Limit your response to 500 words and 8 images.

Answer:

1. Data from MC2 shows that several sensors parked on the Wilson Forest highway since 18:30 9th, trying to get into the area. However, they failed to do so and turned back after a long wait until 8:40 10th. Data from MC3 confirms that the highway was closed again at 19:30 9th, which could possibly predict the traffic on each bridge or highway.
2. MC1’s damage reports can predict people’s asking for help on MC3’s data. For example, when water pipeline's broken records arrived, people in certain areas’ complain would show up in MC3’s.

Q4

The data for the individual mini-challenges can be analyzed either as a static collection or as a dynamic stream of data, as it would occur in a real emergency. Were you able to bring analysis on multiple data streams together for the grand challenge using the same analytic environment? Describe how having the data together in one environment, or not, affected your analysis of the grand challenge. Limit your response to 500 words and 10 images.

Answer:

As three challenges all represent the value tendency over time, data stream would probably bring no great changes to our systems. MC2 uses statistical information to classify different data patterns and would have to wait for a relatively short period of delay before the data accumulated and thus gets reasonable statistical values. However, the delay only happens for latest  data’s patterns distinguishing, and this will not place great influence on our systems.

reliability (color) change with the data stream

different colors represent different patterns

the words stream which can support data stream