Lucas Building Data
Team
• Rajesh Ilango
• Preetha Manohar
• Kalyan Posani
• Ajith Ramachandra
Goal
Analyze the usage pattern to improve the utilization of
building capacity.
•Faculty Rooms
•Conference Rooms
•Classrooms
•Other rooms are 2 Dept. Suites, 2 Staff Room, 1 Common
room and 1 Break room
Questions to Answer
Frequency Of Usage:
•Which rooms are being used the most/least
• Quarterly Analysis
Usage Patterns:
•What days of the week
• Daily Usage
• Weekday vs Weekend
•What time of the day
Rawdata
User Access Data
Masked Users
Different Rooms Data
Polished data
Factors Considered
• Only Faculty usage for faculty rooms
• Only Granted Access
• Considered data from May 2014 to Feb 2015
• Plotted Fall 2014 and Winter 2015 quarterly data
• Faculty Offices, Conference Rooms and Classrooms
• Duplicate entries by a single user on a same day for the
faculty offices were not considered
Faculty Offices:
Usage – High to low
Usage pattern :
• Power law distribution
Take Away:
• Share the faculty offices and vice-versa.
• Better Allocation Strategy
Faculty Offices:
Day-Wise and Per Quarter Analysis
Usage pattern :
• Mon-Thu heavy usage as
expected
• Usage varies drastically by
day and quarter
• Some rooms not used at
all on some days
Take Away:
• Explore the possibility of
more week end classes to
avoid conflict with other
classes
Conference Rooms: (By faculty)
Day-Wise and Per Quarter Analysis
Usage pattern :
• Data reflects very limited
usage
Conference Rooms: (All)
Day-Wise and Per Quarter Analysis
Usage pattern :
• Mostly used by TAs
Conference Rooms : Usage
Usage pattern :
• Heavier usage by non-faculty. e.g.: TAs
• About half the conference rooms were not used at all
Take Away:
• Re-model Conference rooms into Collaboration rooms,
giving preference of usage to Business school students
Classrooms : Usage
Usage Pattern
• Very limited and unclear data
• Doesn’t reflect reality
Take Away
• Existing data is not sufficient to
identify how many classes can
be accommodated in a day.
Interactive plots: iplots
•Interaction between
• Different Room Types
• Day of Week
• Time of the Day
• Weekend vs Weekday
• Monthly Analysis
Enhancements & Limitations
•No data is useless
•Never neglect data
•Limitations
•No Exit Data
•Analysis limited to Faculty usage.
Data Analysis
•Analysis
• Lubridate
•stringr
•Plots
• ggplot2
•iplots
16
Data science project

Data science project

  • 1.
    Lucas Building Data Team •Rajesh Ilango • Preetha Manohar • Kalyan Posani • Ajith Ramachandra
  • 2.
    Goal Analyze the usagepattern to improve the utilization of building capacity. •Faculty Rooms •Conference Rooms •Classrooms •Other rooms are 2 Dept. Suites, 2 Staff Room, 1 Common room and 1 Break room
  • 3.
    Questions to Answer FrequencyOf Usage: •Which rooms are being used the most/least • Quarterly Analysis Usage Patterns: •What days of the week • Daily Usage • Weekday vs Weekend •What time of the day
  • 4.
    Rawdata User Access Data MaskedUsers Different Rooms Data
  • 5.
  • 6.
    Factors Considered • OnlyFaculty usage for faculty rooms • Only Granted Access • Considered data from May 2014 to Feb 2015 • Plotted Fall 2014 and Winter 2015 quarterly data • Faculty Offices, Conference Rooms and Classrooms • Duplicate entries by a single user on a same day for the faculty offices were not considered
  • 7.
    Faculty Offices: Usage –High to low Usage pattern : • Power law distribution Take Away: • Share the faculty offices and vice-versa. • Better Allocation Strategy
  • 8.
    Faculty Offices: Day-Wise andPer Quarter Analysis Usage pattern : • Mon-Thu heavy usage as expected • Usage varies drastically by day and quarter • Some rooms not used at all on some days Take Away: • Explore the possibility of more week end classes to avoid conflict with other classes
  • 9.
    Conference Rooms: (Byfaculty) Day-Wise and Per Quarter Analysis Usage pattern : • Data reflects very limited usage
  • 10.
    Conference Rooms: (All) Day-Wiseand Per Quarter Analysis Usage pattern : • Mostly used by TAs
  • 11.
    Conference Rooms :Usage Usage pattern : • Heavier usage by non-faculty. e.g.: TAs • About half the conference rooms were not used at all Take Away: • Re-model Conference rooms into Collaboration rooms, giving preference of usage to Business school students
  • 12.
    Classrooms : Usage UsagePattern • Very limited and unclear data • Doesn’t reflect reality Take Away • Existing data is not sufficient to identify how many classes can be accommodated in a day.
  • 13.
    Interactive plots: iplots •Interactionbetween • Different Room Types • Day of Week • Time of the Day • Weekend vs Weekday • Monthly Analysis
  • 14.
    Enhancements & Limitations •Nodata is useless •Never neglect data •Limitations •No Exit Data •Analysis limited to Faculty usage.
  • 15.
  • 16.