LocationAware Business Rules

1
Location-Aware Business Rules Events for
Real-Time Sensor Data

• Barry J. Glick
• CEO
• ObjectFX

2
Defining the Need

• Flood of incoming data from sensors,
  communications, transactions, etc
• Location and time conditions help determine
  relevance / significance / needed actions
• Knowledge in form of rules, human interaction,
  etc must be applied
• Real-time data in context physical environment,
  historical trends, knowledge of phenomena
• Organizations benefit from make effective use of
  real-time sensor data for decision-making,
  operations control, monitoring, etc.

3
Typical Sensor Network Process
Value of observable parameters at a location A at
time t
Estimate key parameters at a given location (same
as or near A)
Detect occurrence of event or body of interest
Trigger decision Take action
Classify an object or situation
Track monitor t t1
First Analysis et al, 2004
4
Smart Sensor Network Designthe Role of Location
Design Layers
Decision
Agents, management, distribution, fusion,
storage, parsing, etc.
Information
Locating, routing, tracking, etc
Action
Location
Transport, networking, protocol, topology, etc
Communication
Sensor
Sensor design, materials, size energy
consumption, lifetime, environmental, etc
Environment - Objects
First Analysis et al, 2004
5
Division of Labor in Sensor Data Exploitation for
Decision-Making
Information Bandwidth
Complexity / Knowledge Level High Mid Low
lt101 / min.
102 / min.
gt103 / min.
6
The Value of Location (and Time)

• Filtering and tracking eg.
• Apply movement / velocity thresholds
• Apply density threshold
• Interpolation apply threshold to non-sensed
  location (and time)
• Rules and events eg
• Apply geo-fence conditions (simple to complex) w/
  enterprise data context
• Human interaction eg
• Complex pattern / event recognition
• Decision based on many conditions,
  non-quantitative factors, etc

7
Spatial Processing Requirements

• Filtering tracking
• Position determination, tagging, distance speed
  calculation, interpolation
• Spatio-temporal rules events
• Full range of spatial operators, spatial rules
  management, spatio-temporal pattern analysis
• Human interaction (decision support system)
• Effective dynamic spatial visualization, access
  to full range of spatial decision aids,
  multi-source data integration (including external
  data), etc

8
Characteristics of a Rule-Based System

• Don't specify how to carry out a task
• Encapsulate knowledge or rules about what should
  be done in different circumstances.  
• Control logic is separated from the rules.  
• Rules are usually defined in a declarative,
  human-readable syntax.
• Can be plugged into the application without
  changing any source code.  
• The control logic is embedded in the rule engine
  which decides which rules are evaluated and in
  what order.

9
Spatial Rules and Events

• Provides geospatial monitoring and event
  notification of spatial and location information
• Support for entry, exit, inside, outside,
  distance, density, thresholds, and custom rules
  creation
• Supports actions and notifications through email,
  JMS and custom event creation
• Supports both Real-Time and Retrospective
  analysis
• Built in spatial GUI for easily creating rules on
  spatial points, lines, polylines, polygons, etc
• Applications include data monitoring, anomaly
  detection, change detection, data filtering, etc

10
Spatial Rules and Events Transportation Example
11
Tracking and Business RulesAutomotive Industry
Example

• Combine RFID and business rules to automate
  workflow in vehicle processing centers and rental
  car companies
• Active RFID tag assigned to every vehicle as it
  rolls off the assembly line and is linked to a
  VIN (vehicle identification number).
• Active tags, as opposed to passive tags, have a
  range of approximately 1,000 feet and are
  accurate within a radius of 10 feet.
• Access points distributed around the processing
  center capture the location data. 
• To avoid the tidal wave of data generated by
  RFID, rules based on business process alert
  managers to exceptions and ignore expected data
• Business rules are used to create a processing
  hierarchy based on the business value of the
  products
• The system tracks trailers of inventory around
  yards
• if the trailer has an appointment to be at a
  dock door at a certain time and doesn't arrive,
  it triggers an alert

12
Dynamic Spatial Visualization

• Clear need for innovative methods to provide
  effective graphical environment for depicting
• Complex business situations with many data
  sources / types
• Change over time highlighting anomalies, changes
  in trends, thresholds, etc
• Map and non-map spatial environments
• In-buildings, rail yards, ports, campuses
  (geometry worlds)
• Networks (topology worlds)
• Multi-dimensions, schematic views
• Highly interactive and intuitive

13
Network Visualization Monitoring

• Real-time Network Visualization
• Fault monitoring
• Dynamic Network provisioning
• Locational views
• Improved customer service by increasing response
  time and avoiding systems overloads

14
Military Requirements Example

• Joint Vision 2010 broad use of IT to create
  decision superiority by providing the ability
  to collect, fuse, process and disseminate
  uninterrupted flow of information and C2
• Extensive use of sensors in real-time networks
• Eg program element Dynamic Tactical Defense
• Use existing national/theater intell/surveillance/
  recon sources
• Dynamically task unattended ground sensors,
  unmanned air vehicle sensors, and HUMINT
• Fuse data from sensors with ISR data from all
  sources to enable continuous estimation of target
  location, identity and activity
• Close the loop between sensor management and
  fusion
• Clear role for spatial / temporal filtering,
  rules and events, and visualization for human
  interaction

15
Space-Time Visualization Intelligence and
Military Examples
Intelligence data fusion bringing together
sensor, historical, and other collected data
Operational picture of the battlefield based on
sensor data in multi-source data context
16
Integration Issues

• Because location-based tasks are a subset of the
  complete sensor information management process,
  and needed in many stages, strong integration is
  critical
• Sensor data and other source data ingest
• Where should first stage spatio-temporal
  processing occur?
• Middleware integration complete set of spatial
  and space-time operators and spatial rule engine
  available thru application server
• Forward integration for operational data,
  business process knowledge etc
• Integration with relevant enterprise application
  e.g, supply chain mgmt, CRM, etc
• Web services-based integration issues
• Application development tools integration with
  common studio products

17
RFID Integration Example
Location-enabling benefits
18
(No Transcript)
19
Justifying the Investment in Location Technology

• For organizations that manage and control assets
  in a complex operating environment
• Location technology investment has payback based
  on benefits gained from each of the 3 components
• Helping to intelligently reduce data volumes
  entering the enterprise IT environment and not
  missing the critical situations that need
  attention
• Extending business rules and events capability to
  include spatial and spatio-temporal rules to
  better model key business goals, constraints, etc
• Providing a human-machine environment for
  effective communication of the real-time
  situation of the organization and allowing
  personnel to apply human intelligence in a
  spatial context.
• Leading to better, faster actions and
  decision-making for more efficient asset
  management and control
• Enhanced effectiveness in SCM, military
  operations, FFA, and other enterprise
  applications in which location counts

20
Conclusions

• Clear value for spatio-temporal technology in
  supporting a knowledge-driven real-time
  enterprise
• Manageable data volumes, focus on critical
  information, enforce business rules, effective
  understanding / portrayal of the real-time
  situation
• Multiple levels of integration needed
• Not a single stovepipe application
• Early stage of market
• THANK YOU!
• Email barry.glick_at_objectfx.com