Supporting decision making in security forces command center at - - PowerPoint PPT Presentation

▶
supporting decision making in security forces command
SMART_READER_LITE
LIVE PREVIEW

Supporting decision making in security forces command center at - - PowerPoint PPT Presentation

Dec 29, 2023 19 likes •441 views

Supporting decision making in security forces command center at large-scale events via video-based situation monitoring and base data supply Armin Kfler 1 , Viktoria Pammer-Schindler 2 , Alexander Almer 1 , Thomas Schnabel 1 1 JOANNEUM

slide-1
SLIDE 1

1

Armin Köfler

Supporting decision making in security forces’ command center at large-scale events via video-based situation monitoring and base data supply

Armin Köfler1, Viktoria Pammer-Schindler2, Alexander Almer1, Thomas Schnabel1

1 JOANNEUM RESEARCH Forschungsgesellschaft mbH, Steyrergasse 17, 8010 Graz, Austria 2 Graz University of Technology, Know-Center GmbH, Inffeldgasse 13, 8010 Graz, Austria

Graz, 12.10.2017

slide-2
SLIDE 2

2

Armin Köfler

Content

  • Motivation
  • Approach / Timeline
  • Prototypes / Field studies
  • Research Questions
  • Current Situation & Outlook
slide-3
SLIDE 3

3

Armin Köfler

Motivation

  • Major events are popular and attract huge crowds
  • Carrying out the event in a safe manner is indispensable
  • All involved organizations have a need for situation information
  • Heads of security need information that is exact, current and unambiguous
  • Prevention and improved reaction on hazards through situation visualization
  • Project scope: KIRAS – Security Research: Project MONITOR
slide-4
SLIDE 4

4

Armin Köfler

Approach (1/3)

  • Design Case Study
  • Gaining domain knowledge
  • Iterative process of prototype improvement and assessments by domain users
  • Cognitive Walkthrough
  • Inspection of the usability aspects of prototypes
slide-5
SLIDE 5

5

Armin Köfler

Approach / Timeline (2/3)

1. Rough objective derived from project MONITOR 2. Domain-specific research work 3. Meeting with fire fighting experts (FF) 4. Telephone interviews with FF and Security Chief 5. Discussion of derived features 6. Implementation of prototype P1 after feature selection 7. Feature Demonstration (Workshop)

1 2 3 4 5 6 7

slide-6
SLIDE 6

6

Armin Köfler

Approach / Timeline (3/3)

8. Field study “Nova Rock” 9. Field study “Donauinselfest”

  • 10. Integration of newly recognized features in prototype P2
  • 11. Intermediate feedback during “Frequency Festival”
  • 12. Finalization of feature implementation
  • 13. Workshop

8 9 10 11 12 13

slide-7
SLIDE 7

7

Armin Köfler

Prototype P1

  • A GIS as management support system
  • Desktop application based on C#
  • Focus on presenting current situation based on current data
  • Crowd representation derived from aerial images
  • Visualization and Interaction by using synthetic data
  • GUI Interaction oriented on widely known standard software
slide-8
SLIDE 8

8

Armin Köfler

P1 – Data source ARGUS 1/2

  • Prosilica GE 6600 – RGB camera (29 MPx)
  • INFTRATEC thermal camera (TIR)
  • Novatel GPS, IMAR/FSAS IMU
  • Line of Sight (LoS) Communication
slide-9
SLIDE 9

9

Armin Köfler

P1 – Data source – ARGUS 2/2

  • Example ARGUS RGB with 7cm

ground resolution (600m AGL)

  • Aim is to extract
  • Positions of persons
  • Densities of crowds
  • Aim is a geographically referenced

situation layer for operator

slide-10
SLIDE 10

10

Armin Köfler

P1 – Crowd Visualization

  • Crowds are Heatmap Cells
  • Data basis: Raster 1x1m
  • Each raster cell contains
  • Density (#Persons / m2)
  • Confidence of density
  • Site visually partitioned into
  • ≥ 1 P/m2  colored region
  • < 1 P/m2  transparent region
  • confidence < threshold  Fog of war
slide-11
SLIDE 11

11

Armin Köfler

P1 – Crowd Interaction 1/2

  • Heatmap Cell interpreted as unit
  • Features meta data
slide-12
SLIDE 12

12

Armin Köfler

P1 – Crowd Interaction 2/2

  • Context menu commands / functions
  • CMD: “Estimate cell behavior”
  • FUN: “Remind in 5 minutes”
  • FUN: “Set cell behavior manually”
slide-13
SLIDE 13

13

Armin Köfler

P1 – Snipping Tool

  • Requested from firefighter context
  • Aims at efficient information distribution
slide-14
SLIDE 14

14

Armin Köfler

P1 – Experts workshop – Lessons learned

  • Concept of heatmap cells is deemed applicable
  • Color schema was not favored, ZAMG schema proposed
  • Sending security personnel to cell useful, routing support not wanted
  • Deferred reminding of crowd cells is endorsed
  • Displaying meta information is essential
  • The common site plan is mandatory as background layer (no OSM, …)
  • Snipping Tool is appreciated as additional source for event logs

(G Y O R)

slide-15
SLIDE 15

15

Armin Köfler

Field Study Nova Rock

  • Separate containers for Security and Emergency Services
  • Interaction only via TETRA, radio und cell phones
  • Video-Streams shared between security und police Forces
  • Used operating resources based on paper
slide-16
SLIDE 16

16

Armin Köfler

Field Study Donauinselfest

  • Shared container for Security and Emergency Services
  • Focus on verbal communication between organizations
  • Only ambulance service provides web interface
  • Used operating resources based again on paper
slide-17
SLIDE 17

17

Armin Köfler

Field Studies – Lessons Learned

  • Master data and event log are paper based
  • Bringing the spatial context to mind is cumbersome during calls
  • (Only) verbal communication not reliable due to its volatile nature
  • Inspiration gained by observing specific incident:
  • Need for reduction of mistakes based on usage of non-volatile communication
  • Need for quick landmarks envisionment during radio communication
slide-18
SLIDE 18

18

Armin Köfler

Prototype P2 - Features

  • Focus shift to presentation of master data
  • Efficient finding of places and reference points
  • Support of swift envisionment of state and location during radio call
  • Persistent situation information readily apparent for all of deciding staff
slide-19
SLIDE 19

19

Armin Köfler

P2 – Overview

Interaction Window (Primary Screen) Presentation Window (Beamer, TV)

slide-20
SLIDE 20

20

Armin Köfler

P2 – Interaction Window – Map 1/2

  • Map Grid defined by
  • 3 vertices in WGS84
  • Edge length of cell in meter
  • Mouse position is highlighted in grid
  • Optional mouse label shows grid name
slide-21
SLIDE 21

21

Armin Köfler

P2 – Interaction Window – Map 2/2

  • (Food) booths
  • Extracted from official site plan of securing organizations
  • Labeled with booth number
  • Security personnel
  • Fixed locations based on synthetic data (mockup)
  • Each security person has a number visible on vest
  • Some have a radio at hand
  • Mouse position highlights objects within certain distance

by an orange ring

slide-22
SLIDE 22

22

Armin Köfler

P2 – Interaction Window – Find & Search Box

  • Canvas for mouse context information
  • Text search by
  • Name of Grid Cell
  • Name of all described point objects
  • Number of booth, Number of security person, …
  • Focus map on entity by hitting <Enter>
  • Iterate focus by repeating <Enter> on multiple eponymous objects
  • Red font signals no more matches possible
  • Example for Tool “Find nearest security with radio”:
slide-23
SLIDE 23

23

Armin Köfler

P2 – Presentation Window – Overview

  • Presents a situation to the whole head staff as nonvolatile information
  • Presents situations as
  • Screenshot von map view including “Find & Search Box”
  • Timestamp of creation in tab caption
  • Elapsed time printed over screenshot permanently updated
  • Free-hand annotations per touch and mouse
  • Multiple situations arranged as tabs captioned by absolute time
slide-24
SLIDE 24

24

Armin Köfler

P2 – Presentation Window – Situation Example 1/2

  • Newly captured situation shows
  • Timestamp in tab title
  • Relevant Content of Interaction Window
  • Elapsed time on top
slide-25
SLIDE 25

25

Armin Köfler

P2 – Presentation Window – Situation Example 2/2

  • Annotated situation showing
  • Timestamp in tab title
  • Relevant Content of Interaction Window
  • Elapsed time on top
  • Annotations
  • Sketches
slide-26
SLIDE 26

26

Armin Köfler

P2 – Expert workshop – Lessons learned

  • Head of security names the Presentation Window valuable
  • Adaption of used hotkeys advised
  • Search Box is rated an improvement in case processing
  • Highlighting master data objects on map depending on mouse distance is

deemed valuable for bringing spatial context to mind

  • Reenacted situation with DIF-Operator allowed swifter detection of mistake
  • Expanding the kinds of master data is desired (Camera locations, ...)
slide-27
SLIDE 27

27

Armin Köfler

Research issue 1/2

  • Does the presentation of a current situation as an interactive digital map

improve the situational awareness of the security operator?

Seeking reference objects speeds up spatial envisionment Digital form is superior to paper form (but keep redundant!) The visualization encourages reflection

slide-28
SLIDE 28

28

Armin Köfler

Research issue 2/2

  • How can a situation made to be known to all relevant headquarter staff?

How can mistakes be mitigated / reduced?

Open presentation of situations is key It creates a culture of obligation to fetch information for anyone Discrepancy in presented information shall motivate to actively discuss unclear or

contradictions

slide-29
SLIDE 29

29

Armin Köfler

Current Situation & Outlook

  • Combination of the core features of prototypes into one
  • (Utilization of prototype at Frequency 2017)
  • Expert workshop by the end of October for broader feedback
slide-30
SLIDE 30

30

Armin Köfler

Discussion of further development

slide-31
SLIDE 31

31

Armin Köfler

Thank you for you attention

slide-32
SLIDE 32

32

Armin Köfler

ANNEX

slide-33
SLIDE 33

33

Armin Köfler

Control-Loop of leading

slide-34
SLIDE 34

34

Armin Köfler

Thresholds of crowd densities as heatmap

  • (Hillen/Meynberg/Höfle 2015) suggest that walking speed is indirectly

proportinal to crowd density

  • Ableitung von Dichte aus Gehgeschwindigkeit möglich
  • Typische Gehgeschwindigkeit 1,32 bis 1,51 m/s
  • Kontext relevant (Tätigkeit, Alter, Geschlecht)

Schwellwert der Dichte [p/m2] Eigenschaft 0,43 Normale Gehgeschwindigkeit 2 Reduzierte Gehgeschwindigkeit 3,57 Unbeabsichtigtes Rempeln 5,55 Potentiell gefährdende Drücke 7,1 Stillstand

slide-35
SLIDE 35

35

Armin Köfler

Existing GIS based solutions

  • Intelli R.4C - Stabs- und Führungssystem für Großschadenslagen
  • Euro DMS LUMIS - digitale Einsatzführungsunterstützung (FF Graz, Castor)
  • Kappacher ELDIS - langfristig laufendes Leitstandsystem für Polizei und FW
  • PointR - Indoor-Positionierung und -Navigation mit Analyse der

Bewegungsmuster in Form von Heatmaps

slide-36
SLIDE 36

36

Armin Köfler

Photos Nova Rock / DIF

slide-37
SLIDE 37

37

Armin Köfler

P2 – Interaction Window – Features

  • Search by keyboard, mouse or touch
  • Map content: Map grid, Booths, Security personal
  • Find & Search Box
  • Visualization Mouse Context
  • Text Search and Navigation
  • Minimap
  • Toolbox for Orientation and Queries
slide-38
SLIDE 38

38

Armin Köfler

P2 – Interaction Window – Minimap

  • Pre-defined global borders
  • Red frame shows current main view
  • Click for navigation
slide-39
SLIDE 39

39

Armin Köfler

P2 – Interaction Window – Toolbox

  • Navigation
  • Captions on map
  • Displaying labels of map objects
  • Displaying map grid caption at mouse pointer
  • Queries
  • Example: „Seek next guard with radio at click location“
  • Partial screenshots via Snipping Tool
  • Copy to clipboard for arbitrary usage
  • Auto-create new Word document with inserted image
slide-40
SLIDE 40

40

Armin Köfler

References

  • Hillen, Florian/Meynberg, Oliver/Höfle, Bernhard (2015): Routing in Dense

Human Crowds Using Smartphone Movement Data and Optical Aerial Imagery. In: ISPRS International Journal of Geo-Information 4, S. 974–988.

  • Image DIF von https://www.facebook.com/WienerPolizei.at/?fref=photo
  • „Control-Loop of lead “ translated from ÖNORM S 2304
slide-41
SLIDE 41

41

Armin Köfler

P2 – Cognitive Walkthrough (CW) 1/2

  • CW is used for discovering shortcomings in GUIs
  • Test subjects are developers not affiliated with the app’s domain
  • Given tasks are divided into single steps
  • Every step is analyzed by four questions:

Step fulfills goal? Is action visible? Can action be recognized amongst others? Can user understand the system feedback?

slide-42
SLIDE 42

42

Armin Köfler

P2 – Cognitive Walkthrough 2/2

  • Logged using screencasts and forms
  • Various problems were found and fixed
  • Most relevant improvements:
  • Temporary markers are not removed after 5 seconds
  • Text search seeks potential finds after each key hit
  • Existence of eponymous objects is stressed