A Study of Firefighting in the Coming Age of Ubiquitous Computing

1
A Study of Firefighting in the Coming Age of
Ubiquitous Computing
G r o u p f o r User Interface Research
University of California Berkeley

• Xiaodong Jiang
• Leila A. Takayama
• Jason I. Hong
• James A. Landay

2
Research Motivation

• Emergencies are a fact of life
• 1,755,000 fires in the United States in 1998
• 4000 deaths, 22000 injuries, 100 firefighter
  deaths / year
• 9 billion in property losses / year
• Difficult to make coordinated decisions under
  stress
• Assessment, tracking, communication, and planning
• "Firefighting is making a lot of decisions on
  little information"
• Improvements can save lives and minimize damage
• Every second counts

3
Research Motivation

• Relatively unexplored domain in HCI
• Very far away from desktop computing
• High-stress and chaotic
• Sensor nets
• Small cheap sensors for location,
  identity, temperature,
  humidity
• Significantly pushes scalability and usability
  issues for ubicomp apps

4
Overall Approach (1 of 2)

• Field studies of Firefighters
• Understand existing tools, processes, language,
  structure of emergency responders
• Understand the problems and constraints they have
• What we learned
• Incident Command System for managing and
  coordinating resources
• Accountability
• Assessment
• Communication
• Inside the Fire

5
Overall Approach (2 of 2)

• Iterative development of lo-fi prototypes
• Understand what kinds of sensor information
  useful, who needs to know it, and how to present
  it
• Prototype an electronic board for relaying sensor
  info to a specific role, the Incident Commander
• Prioritize for "bang for buck" deployment
• What we learned
• Location is the most important info
• Also wanted support for managing resources
• Difficult to scale for large incidents

6
Outline

• Motivation
• Field Studies
• Low-Fi Prototypes

7
Field Studies

• Field studies
• Observed and interviewed participants in "normal"
  work environment
• Four months, three depts
• One field exercise
• Two emergency calls
• Participants
• 1 Assistant Chief
• 5 Battalion Chiefs
• 2 Captains, 2 Engineers
• Many Firefighters

8
Firefighter Organization

• Basic unit of organization is the Company
• "Any piece of equipment having a full complement
  of personnel"
• Engine, Truck (Ladder), Brush, HazMat
• 1 Captain, 1 Engineer or Driver, 1 Firefighters
• Battalions are a collection of companies

9
Field StudiesOrganization
10
Organization in Larger Incidents

• Companies are organized hierarchically
• Divisions are for geographic regions
• Ex. North or Third Floor
• Groups are for specific functions
• Ex. Ventilation, Rescue, Rapid Intervention Team
• Command Post
• Officers and staff managing overall response
• Planning, Operations, Logistics
• Person in overall command is the Incident
  Commander (IC)

11
(No Transcript)
12
(No Transcript)
13
Incident Command System (ICS)

• Used by many local, state, and federal agencies
• "Unified command, common terminology,
  comprehensive resource management, and manageable
  span of control"
• Five roles
• Command -gt Strategic plan
• Operations -gt Tactical operations
• Planning -gt Maps, weather reports
• Logistics -gt Getting supplies
• Administration -gt Finances

14
Example ICS Forms

• ICS supported by many forms
• ICS form 201
• Help visualize and keep track of situation,
  communicate with others

15
(No Transcript)
16
(No Transcript)
17
Scenario Single Story House Fire

• 911 Dispatcher notifies nearest Fire Station
• Firefighters in first engine size up the
  situation
• Ex. Layout of building, scope of fire, nearest
  hydrants
• Engineer sets up hose lines
• Highest ranking officer assumes role of IC

18
Scenario Single Story House Fire

• Battalion Chief arrives
• Assumes role of IC
• Might also assume Operations and Logistics
• Delegates in larger incidents
• Gets 30-sec assessment from previous IC
• What resources do you have?
• Who is here, where are they?
• Status of fire?
• What resources are needed?

19
Passports
20
Scenario Single Story House Fire
21
Scenario Single Story House Fire

• Figures out a plan
• Offensive / Defensive
• Call for more resources
• Second alarm, Third alarm, etc
• Once the fire is extinguished, releases resources

22
Key Findings Accountability

• Accurate count of resources personnel
• Rapid notification of immediate dangers
• Some approaches
• Two-in two-out
• Roll calls
• Passports
• PASS
• Problems
• Chaotic, difficult to get good info
• Situations change quickly

23
Key Findings Assessment

• Sizing up the situation correctly
• Scope of fire, hidden fires, floorplans, dangers
• Some approaches
• Prevention (annual inspections, drills)
• Collection of info beforehand
• Material Safety Data Sheets
• Floorplans
• Firefighters on scene radioing back info
• Problems
• Data out of date
• Difficult to find right info
• Difficult to get right info

24
Key Findings Communication

• Coordinating responders
• Some approaches
• Face-to-face
• Radio communication
• Problems
• Noise intensity
• Congestion
• Radio dead zones
• Missed orders
• Missed abandons

25
Key Findings Inside the Fire

• Carrying 40 lbs of equipment
• Jacket, SCBA, Axe, etc
• Often can't see due to smoke
• Crawling on ground
• Stay near hose lines, guide ropes, or right-hand
  searches
• PASS system
• Panic button, motion sensor
• Most are audio only

26
Key Findings Inside the Fire

• Flashovers
• Simultaneous and sudden ignition
• New dangers due to equipment
• Backdrafts
• Oxygen starved fire gets oxygen
• Hidden fires
• Structural Collapse
• Personal Hazards
• Getting lost, running out of oxygen,
  disorientation

27
Outline

• Motivation
• Field Studies
• Low-Fi Prototypes

28
Low-fidelity Prototypes

• Initially focus on the IC
• Three low-fi prototypes of electronic board
• Understand what kinds of sensor info are useful,
  how to present it

29
(No Transcript)
30
Prototype 1

• Pros
• Floor plans very useful
• Tracking individual firefighters useful
• Good for small incidents
• Cons
• Unsure if could get sensor info
• Some info useful but too detailed for large
  incidents
• Need better support for managing resources
• History not very useful

31
(No Transcript)
32
Prototype 2

• Pros
• Liked the ICS greaseboard metaphor
• Liked having overview map and local map
• Sensor data about companies kept on the edge
• Cons
• ICS hierarchy not often used, wastes a lot of
  space
• Hard to see important info when needed

33
(No Transcript)
34
Prototype 3

• Pros
• "Resource-task-location" very well received
• Tracking of how long on duty also well received
• Notification of critical situations better
• Scales better for larger incidents
• Cons
• Mixes Command, Planning, Ops
• Concerns about cost, implementation and
  reliability

35
Summary

• Incident Command System
• Accountability, Assessment, Communication,
  Inside the Fire
• Three low-fidelity prototypes
• Location very useful
• Originally wanted sensor-based apps, but basic
  resource management very useful
• Scale is still a very difficult problem
• Also questions of implementation (reliability)

36
Thanks to Berkeley Fire Dept El Cerrito Fire
Dept Alameda Fire Dept NSF ITR CITRIS Nick,
the camera man
G r o u p f o r User Interface Research
University of California Berkeley

• Xiaodong Jiang
• Leila A. Takayama
• Jason I. Hong
• James A. Landay
• http//guir.berkeley.edu/emergency