![R equirement Analysis for Mobile Information Exchange in the Police Using a Time-Aware and Agent Based Approach Raul Savimaa Tallinn University of Technology [email_address]](https://image.slidesharecdn.com/requirement-analysis-for-mobile-information-exchange-in-the-police-using-a-timeaware-and-agent-based-approach-7118/85/Requirement-analysis-for-mobile-information-exchange-in-the-police-using-a-time-aware-and-agent-based-approach-1-320.jpg)
Requirement analysis for mobile information exchange in the police using a time-aware and agent based approach
- 1. R equirement Analysis for Mobile Information Exchange in the Police Using a Time-Aware and Agent Based Approach Raul Savimaa Tallinn University of Technology [email_address]
- 2. Contents of the presentation Research domain Requirements for p rocess and b ehaviour m odels Existing m odelling m ethods Possible suitable a pproaches Review of t he applied novel m ethodology T he c ase s tudy: requirement analysis for m obile i nformation e xchange in the p olice Conclusions
- 3. Research domain Research concentrates on multifunctional human organisations that Operate in a dynamic environment Behaviour is time-sensitive, response to emerging situations must be made in minutes and seconds Decisions may have serious impacts A pre-defined set of action rules exists Many decisions are left to structural units (employees), but their specific goals and priorities are not analysed
- 4. General r equirements for p rocess and b ehaviour m odels Encapsulation of most important work processes and organisational aspects Representation of interactions and co-operation mechanisms between actors Straightforward interpretation of the modelling results to everyday practice Enable persistent adjustment of the model
- 5. Specific requirements for modelling time-sensitive behaviour Modelling all temporal criteria of processes and interactions, not just only ordering of activities Modelling and comparison of the planned and emerging behaviours Modelling of all essential aspects that influence organisation behaviour and behaviour of actors, their choices and preferences
- 6. Existing m odelling m ethods Planning and assessment approaches: Gantt chart, PERT/CPS, SWOT, PEST Methods for practical evaluation of organisational performance: Balanced Scorecard, Best Value Workflow management and enterprise modelling approaches: Workflow Reference Model, EKD
- 7. Why existing conventional solutions are not sufficient Concentrate on processes and information flows only Do not consider time issues important Do not consider motivations in relation with choices and processes
- 8. Chosen approach is based on Models of interactive computing (Wegner, 1998); Emergent behaviour research (Carley and Prietula, 1998; Alatalo, 2002) UML (Booch, Rumbaugh and Jacobson, 1999) Analysis of timing criteria – the Q-model (Motus and Rodd, 1994) Multi-agent systems (e.g. Ferber, 1999; Wooldridge, 2002)
- 9. The s uggested novel m ethodology Bases on integration of three approaches: Work processes and actors are described by UML use case, activity, and interaction diagrams Timing parameters and interactions of processes are specified using the Q-model (a model processor for UML) Co-operation and interactions between its actors is specified and simulated as a multi-agent system
- 10. T he c ase s tudy : m obile i nformation e xchange in the p olice Analysis and modelling consisted of three stages E xisting and planned work processes were specified by using UML use case, activity and sequence diagrams Temporal criteria of k ey processes and interactions were modelled as the Q-model diagrams T he suggested system was simulated as a multi-agent system in JADE environment 8 human and artificial actors and their roles were modelled : police officials , information system , communication module, other vehicles and drivers
- 11. Results of t he c ase s tudy The current research project was successful P rocesses were modified according to analysis 230 police vehicles are equipped with a computer and perform online queries into police databases. T he chosen solution supports police work and makes information exchange more effective A pplication of the methodology was important for C orrect analysis of system requirements M odelling and simulation of actors’ behaviour F or choosing suitable technical solutions F or validation of the corresponding information system
- 12. Conclusions T ime-sensitive emergent behaviour of a human organisation that operates in a dynamic environment can be efficiently modelled Theoretically and methodologically, there are not many known attempts of building time-aware proactive models for a human-centred organisation. The new approach has been applied and tested on separate fragments of a large organisation and the results could be considered successful