Concluding Comments

1
Concluding Comments

• Mike Goodchild
• University of California
• Santa Barbara

2
Defining the problem

• Existence of roads
• geometry
• scale
• centerline or double line
• lanes
• Attributes of roads
• for georeferencing
• for maintenance
• for navigation
• Network topology

3
Defining the problem (2)

• Product-oriented
• QA, updating
• dissemination
• Transaction-based
• master version
• gatekeeping
• institutional arrangements
• business model

4
Data acquisition
Database
Product
User
5
Data modeling

• Enterprise-wide
• an integrated view of the enterprise
• anticipating applications
• Types of data models
• essential
• comprehensive
• standard

6
Accuracy

• Application-dependent
• Often not well-defined
• Truth in labeling
• fitness for use
• for every level of accuracy there exists at least
  one application for which that level is not
  adequate
• benefits and costs
• there is no one product that fits all

7
Context-dependence

• The difficulty of the problem varies spatially
• and temporally and with scale
• trees, cars
• cloud
• geometric complexity of street network
• surface discrimination
• a map of CLEM difficulty

8
The role of the human

• The human will never disappear from the process
• but which human, and where in the process?
• Chicken and egg
• introducing IT to a human process
• where does IT add the most value?
• introducing humans to an IT process
• where does the human add the most value?

9
The role of prior knowledge

• Updating
• new construction
• conflation
• improved geometry
• corrected attributes
• time-dependent attributes
• incidents
• construction
• need an enterprise data model
• a weakness in UNETRANS?

10
The role of prior knowledge (2)

• Predictive modeling
• subdivision plans
• Anticipating change
• locating new subdivisions

11
Data acquisition

• Remote sensing
• universal coverage, open access
• costly, cloud, 99 redundancy, time delays
• airborne vs orbiting
• Probes
• GPS van
• delivery services
• emergency services
• local government

12
Data acquisition schedules

• Transaction-based
• UK OS
• 105 organizations
• potential collaborators
• institutional scale
• Project-based
• trigger factors
• project scale
• deteriorating value
• funding issues

13
Analogous problems

• Line-following
• image processing
• vectorization
• map scanning
• fingerprints

14
New technologies

• Data modeling
• New applications
• LBS
• New sensors
• New probes

15
(No Transcript)
16
Next steps

• Sharing of information
• web-based
• meetings
• Research
• monitoring new technologies
• conflation
• remote sensing
• ???