Entry Name:  "Midd-Andrews-MC2"

VAST Challenge 2016
Mini-Challenge 2

 

 

Team Members:

Replace this list of team members with the names, affiliations, and email addresses of your own team. Indicate the primary point of contact.  Example:

Christopher Andrews, Middlebury College, candrews@middlebury.edu     PRIMARY
Shannon Ovitt, Middlebury College,
smovitt@middlebury.edu

Lily Taub, Middlebury College, ltaub@middlebury.edu

Student Team:  YES (+ 1 professor)

 

Tools Used:

Middguard, a web-based tool building framework developed by C. Andrews and his students in his lab

Valentina Studio

 

Approximately how many hours were spent working on this submission in total?

500 hours

 

May we post your submission in the Visual Analytics Benchmark Repository after VAST Challenge 2016 is complete? YES

 

Video 

 http://www.cs.middlebury.edu/~candrews/media/Midd-Andrews-MC2.wmv

 

 

Questions

MC2.1 – What are the typical patterns visible in the prox card data? What does a typical day look like for GAStech employees?

Limit your response to no more than 6 images and 500 words.

Most employees come in, go to their office and stay there except for meetings, lunch and occasional restroom stop:

Security

Floor one security: 8a-5p, staggered lunch out (12-1 and 1-2). Both visit the deli in morning and at 3:00. After 3:00 convene in 1010

Floor two security: 8a-5p, lunch out 12-1. Spend most of their days in 2170.

Floor three security: 8a-5:30, lunch out 12:30-1:30.

Security has meetings on Tuesdays and Thursdays 8:10a-8:54a. First floor security alternate who comes up for the meeting. Mihalis Vollan takes an early lunch and covers 1010 while Stenig Fusil is at lunch.

../images/security_day.pngIT

                  First shift: 8:00-5:00, lunch 12:00-1:15 in the deli, two meetings a day with engineering in 2700, 10:30a-11:30a and 2:00p-3:00p

                  Second shift: 4:00p-12:00a, meal 8:15-8:45 in the deli, two meetings a night with engineering in 2700, 6:30a-7:30a and 10:00p-11:00p

Engineering

                  First shift: 7:30a-5:00p, lunch 12-1 in the deli, two meetings a day with engineering.

                  Second shift: 4:00p-12:00a, meal 8:00-9:00 in the deli, two meetings a night with engineering

../images/it_engineering_day.png

 

Facilities

                  First shift: 12:00a-7:00a, mostly stay in office, going to 1030 for 2:30-3:30 and 6:00-7:00 every morning

                  Second shift: 7:30a-5:00p, lunch 12-1 in the deli, breaks at 10:30 and 2:30 (they go to the prox zone with the break room)

                  Third shift: 4:00p-12:00a, meal 8:00p-9:00p, breaks at 6:30p and 10:00p

../images/facilities_day.png

Administrators

                  8:00a-5:00p, varies when they take lunch, but it is typically an hour and usually at the deli.

../images/administration_day.png

HR

                  8:00a – between 4:30p and 5:30p, irregular meals, possibly eating in offices, meetings on Wednesdays 2:30-4:15

../images/hr_day.png

Executives

                  7:30a/8:00a – 5:00/6:00, irregular lunches, less constancy in schedules

../images/executive_day.png

 

MC2.2 – Describe up to ten of the most interesting patterns you observe in the building data. Describe what is notable about the pattern and explain what you can about the significance of the pattern.

The temperature and REHEAT Coil power are placed into a wild oscilations in room 3000 shortly before the first major hazium spike. This is notable because the swings are large (32 is quite hot and 8 is quite cold), and it is left uncorrected for the remainder of the period. The fact that this occurs in room 3000 where we have our first major spike in hazium readings cannot be ignored either.

../images/room%203000%20hvac.png

 

Most of the HVAC systems display very regular daily patterns. This is key, because when they are disrupted around the spikes in hazium, the changes stand out and are observable.

../images/supply_fan.png

 

MC2.3Describe up to ten notable anomalies or unusual events you see in the data. Describe when and where the event or anomaly occurs and describe why it is notable. If you have more than ten anomalies to report, prioritize those anomalies that are most likely to represent a danger or serious issue for building operation.

Limit your response to no more than 10 images and 1000 words.

Patrick Young's prox card is stolen

  -  After getting second prox card his first card is only used to enter server room 3440 from the second floor

  - the prox card registers in the server room at 10:20 on the 2nd, 3rd, 8th, and 10th while

    the second prox card shows Young in his office

  -  The mobile prox registers the missing card with lbennet while the static system shows it elsewhere

  - This suggests that Young's prox card was stolen in order to conduct suspicious business in the server room

../images/server_breakin.png

 

../images/stolen_card.png

 

HVAC abnormalities during hazium spikes on June 7th and 8th

  - During the hazium spikes on floors 2 and 3 (June 7th/8th), the following HVAC sensors exhibit unusual behavior:

      - The temperature swings in F2Z2 and F2Z4

      - The VAV Damper Position in the server room oscillates

      - Spike in HVAC Electric Demand Power

      - SYS Supply Fan: Fan Power (Floor 3) becomes lower than usual and oscillates

      - SYS Supply Fan Outlet Temp on Floor 3 spikes, on Floor 2 drops

      - Outdoor airflow fraction spikes and oscillates on Floor 3 and spikes on Floor 2

      - Cooling Coil Power on Floor 3 drops

      - Airloop Inlet Temp on Floor 3 spikes and oscillates, on Floor 2 drops

      -Reheat Coil Power spikes

      - This demonstrates a clear HVAC relationship with rising hazium levels

 

    ../images/HVAC%20response.png 

 

../images/temperature.png

 

 

HVAC abnormalities during Hazium spike on June 11th/12th. While these also seem to accompany the rise in hazium levels, the effects are different.

    - Hazium levels begin to rise June 11th at 12pm, peaking at 6pm, and falling back to minimal levels by June 12th at 6pm

    - Floor 1 and 2 supply Fan power is high until peak and then begins to oscillate, while Floor 3 remains high the whole time

    - Supply Fan Mass Flow Rate is higher than normal on all floors

    - Floor 1 and 2 Outdoor airflow drops

    - Floor 1 and 2 Cooling Coil Power spikes

    - Floor 1 and 2 Air Loop Inlet Mass Flow Rate oscillates

    - Floor 1 and 2 Reheat Damper Position oscillates

    - Floor 1 Mechanical Ventilation Mass floor Rate is high

    -There is abnormal fluctuation in Total Power levels

    - Reheat Coil Power oscillates

    - The breaking of normal HVAC patterns demonstrates a connection between building atmosphere and Hazium levels

 

../images/temperature2.png

../images/reheat_coil.png

 

 

 

Four facilities employees leave their prox cards behind when they leave for the night

  - Dante Coginian leaves his prox card in F2Z1 (potentially in break room) on Tuesday nights (May 31st, June 7th)

  - Effie Davies also leave her prox card in F2Z1 on Tuesday night, May 31st

  - Valeria Morlun leaves her prox card in F2Z1 on Tuesday night, June 7th, while Adan Molrun leaves his in the elevator/stairs (F2Z4) on the same night

 

../images/left_cards.png

 

Lost Prox Cards

  - those who have lost their prox card receive additional prox card IDs that are registered in the building

  - most employees lose their prox cards in the lobby or in their office

  - Jae Unger loses his prox card on June 9th

  -  Lucas Alcazar loses his prox card on June 6th

  - Valeria Molrun loses her prox card on June 3rd

  - Patrick Young (who has one of his prox cards stolen) receives a new prox card on June 2nd

  - Geneviere Florez loses one prox card on the 1st, another on the 2nd, and a final on the 3rd

  - Twana Quirox loses her prox card on the 7th

 

  ../images/lost_prox_cards.png(

 

 

Half of the security employees never show up in the data

    - Dorothea Bartley, Mi Canada, Summer Holiday, Paulita Kemp, Mathias Notaro, Alcide Pherigo, Dong Rhoads, Selma Salvay, Xiomara Sowell, Edvard Vann

 

 

  Coming in on the weekend

    - both Orhan Strum and Mat Bramar come in on Saturday, June 11th (the day of a major hazium spike)

    - Interesting because this is unusual and marks the beginning of a large spike in hazium levels.

 

../images/weekend_hazium.png

 

MC2.4– Describe up to five observed relationships between the proximity card data and building data elements. If you find a causal relationship (for example, a building event or condition leading to personnel behavior changes or personnel activity leading to building operations changes),  describe your discovered cause and effect, the evidence you found to support it, and your level of confidence in your assessment of the relationship.

Limit your response to no more than 10 images and 1000 words.

When meetings occur, we can detect them through upswings in CO2 levels and temperature. The most obvious instance of this occurs in the daily meetings at 10:30a and 2:00p of the joint Engineering and IT departments. We are very confident of this as we can see when the departments go to prox zone F2Z6 and we can see clear bumps in the HVAC levels of 2700 at these points.

../images/cO2_meetings.png

 

Light power and equipment power show when employees arrive and leave. This is most obvious in single room HVAC zones such as the conference rooms and the corner offices. We are highly confident of this.

../images/power_correspondence.png

 

We speculate that increased levels of hazium or related effects to the HVAC system are causing discomfort. Several employees break their patterns around the time of spikes in hazium activity. Rachel Pantanal, who has an office on the third floor, goes home early during the first major spike in hazium levels. Ruschella Haber and Linda Lagos  change their lunch habits, leaving the building for lunch on the day of the first spike (and Linda also leaves the day of the second one). The connection here is weak, but merits further investigation.

../images/leaving_early.png