Towards middleware based situation awareness

1
Towards middleware based situation awareness

• Leo Motus, Merik Meriste ja Jürgo-Sören Preden

2
Distributed teamwork difficulties

• The team comprises autonomously operating, and
  interacting mix of humans, computers and
  software-intensive artefacts
• Each member of the team may have its own
  environment, sub-goals, cognitive capabilities,
  and situations guiding their operation
• Information from the environment and about
  members state comes from different sources, and
  need not be consistent with the goals set for the
  team
• The problem check the consistency of
  information, harmonise its interpretation, and
  achieve common understanding

3
Example of a problem with mixed team

4
The idea for solving the problem
Middleware digital map

5
Situations

• A situation is the aggregate of biological
  psychological, socio-cultural, and environmental
  factors acting on individual, or group of agents
  and may condition their behavioural patterns
• Each autonomous agent may define its own
  situations individual situation awareness
• In a team we need team situation awareness a
  harmonised model for individual situation
  awarenesses and for joint decision-making process
• Team situation awareness results from the
  negotiation process.

6
Steps in the negotiation process

• Agree on the common ontology, information
  presentation form, data structures
• Verify the exchanged messages
• Fuse messages into more abstract messages, if
  necessary
• Achieve common understanding of the actual state
  and course of actions the team situation
  awareness
• Achieve common agreement on actions taken
• Coordinate, or synchronise, the actions
• Collect and analyse the feedback from the actions
  to improve the process

7
Fusing the situations
8
Middleware

9
Architecture of the middleware enables self-X
communication (1)

• External (application oriented) services
• verification and validation of messages and
  information
• transformation and compression into interim
  format
• keeping tracking of the position of agents and
  their access rights
• memorising and using the subscriptions and
  processing capabilities of agents
• delivering the subscribed information within the
  personalised constraints and restrictions

10
Architecture of the middleware enables self-X
communication (2)

• Internal services, implemented by a multi-agent
  digital map system
• handling the interim data format
• provide the base map for the required area
• estimate, maintain and update the position of
  application agents
• partition the map based on the subscription from
  the application agents, e.g. specific areas or
  layers, specific background information, specific
  formats , etc
• memorising interest points of subscribers

11
Conclusions

• Practical impact of mixed teams (humans,
  computers, software-intensive artefacts) on
  functioning of our society increases rapidly.
• Formally such teams can be modelled as proactive,
  ad hoc, and mobile networks that integrate
  networks of proactive pervasive computing
  systems, mobile, proactive smart artefacts, and
  elements of social networks.
• The suggested middleware for such networks
  requires several innovative research results,
  e.g.
• Super-Turing models of computation
• Verification methods for networked systems