Academia.eduAcademia.edu

Figure 1 – uploaded by Victor Callaghan

See full PDFdownloadDownload figure
Figure 1. Example of the IN showing the dependiencies of 5 services in an intelligent computing environment The functionality of a node in a digraph is defined as the number of descendants in the Interaction Network. This characteristic of a node is very important, as it shows the impact of a process on the wider system, in terms of the number of processes whose activity rules could be triggered. Figure 1 provides an example of an Interaction Network in an intelligent environment (the sort of system that might be used in an intelligent building or smart home), showing the dependencies of 5 network services: Sofa Sensor, Light Sensor, MP3 Player, Light, and Word. The light depends on the state of the light sensor, and in the state of the MP3 player. The software application word depends on the state of the light, and in the occupancy of the sofa (sofa sensor). Additionally, and accordingly to our formal definition, each agent A, has been assigned a weight: sofa sensor (High),

Figure 1 Example of the IN showing the dependiencies of 5 services in an intelligent computing environment The functionality of a node in a digraph is defined as the number of descendants in the Interaction Network. This characteristic of a node is very important, as it shows the impact of a process on the wider system, in terms of the number of processes whose activity rules could be triggered. Figure 1 provides an example of an Interaction Network in an intelligent environment (the sort of system that might be used in an intelligent building or smart home), showing the dependencies of 5 network services: Sofa Sensor, Light Sensor, MP3 Player, Light, and Word. The light depends on the state of the light sensor, and in the state of the MP3 player. The software application word depends on the state of the light, and in the occupancy of the sofa (sofa sensor). Additionally, and accordingly to our formal definition, each agent A, has been assigned a weight: sofa sensor (High),

Related Figures (7)

where A = {A,} is the set of agents that stabilize the system. In figure 2 we illustrate the high-level algorithm. One of the goals is minimizing the number of agents locked, in order to have a less-disabled system. The concept of weak and strong coupling cycles has been proven to be valuable tool on analyzing and understanding complex systems [8].
where A = {A,} is the set of agents that stabilize the system. In figure 2 we illustrate the high-level algorithm. One of the goals is minimizing the number of agents locked, in order to have a less-disabled system. The concept of weak and strong coupling cycles has been proven to be valuable tool on analyzing and understanding complex systems [8].
Figure 3. Business organisation & agent organisation as a set of processes
Figure 3. Business organisation & agent organisation as a set of processes
Figure 4. Grid with 64 business processes (nodes) and random inter- connections. The system had 14 cyclic processes. The experiment was implemented using Mathematica® 6, a programming language with powerful tools for quick prototyping. In particular, the package Combinatorica proved extremely useful, as it provided tools for graph theory, graphics and combinatorics. Using the simulator it was possible to control a number of parameters, such as the number of business processes involved, probability of managerial perturbations, generation of random company topologies and random rules of interaction, amongst others. We used the digraph theory in section 2 to find that there were 14 phantom work cycles
Figure 4. Grid with 64 business processes (nodes) and random inter- connections. The system had 14 cyclic processes. The experiment was implemented using Mathematica® 6, a programming language with powerful tools for quick prototyping. In particular, the package Combinatorica proved extremely useful, as it provided tools for graph theory, graphics and combinatorics. Using the simulator it was possible to control a number of parameters, such as the number of business processes involved, probability of managerial perturbations, generation of random company topologies and random rules of interaction, amongst others. We used the digraph theory in section 2 to find that there were 14 phantom work cycles
Figure 5. Dynamics of a business process system with 64 nodes and 81 cyclic processes.
Figure 5. Dynamics of a business process system with 64 nodes and 81 cyclic processes.
Figure 6. Response of a system with 64 nodes and 14 cycles when stabilising action was taken. Using this analysis data, and with some thought by the management, it is possible to stabilise the dynamics of the business processes by carefully disabling or reorganising some cyclic processes. Of course, the decision on which business processes to alter is critical and there is a future opportunity hat by designing the tool to be intelligent we would hope to be able to learn such decisions which, when used in combination with the live lock algorithms would produce a more intelligent business tool. For the example, by way of an example we have applied the disabling vector: {1,1,1,1,1,1,1,1,1,1,0,1,1,1,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,1, 1,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,1,1,1,1} which has stabilised the company cyclic processes (Fig. 6).
Figure 6. Response of a system with 64 nodes and 14 cycles when stabilising action was taken. Using this analysis data, and with some thought by the management, it is possible to stabilise the dynamics of the business processes by carefully disabling or reorganising some cyclic processes. Of course, the decision on which business processes to alter is critical and there is a future opportunity hat by designing the tool to be intelligent we would hope to be able to learn such decisions which, when used in combination with the live lock algorithms would produce a more intelligent business tool. For the example, by way of an example we have applied the disabling vector: {1,1,1,1,1,1,1,1,1,1,0,1,1,1,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,1, 1,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,1,1,1,1} which has stabilised the company cyclic processes (Fig. 6).
Figure 7. This graphs shows cyclic processes for 64 nodes using an MDM
Figure 7. This graphs shows cyclic processes for 64 nodes using an MDM
Figure 8. Density and usability of a system These findings, particularly the destructive and undesirable haracteristics of phantom work cycles, are consistent with omplex interactions observed in intelligent environments [36]. ‘rom Zamudio’s observations, and the findings of our work we surmise that, in business processes, Design-Build-Tes ire necessary for new produc development, and in the cycles case of ntelligent environments cyclic and coupled interactions can rovide specific services to the final user. Extrapola ting the esults of Zamudio’s earlier work suggests that in both domains here is a threshold poin yerformance decreases drama’ © from which the system tically: in the case of in mvironments the system wi 1 self-lock, and hence telligent become musable; in business process redundant innovation cycles can ead to the creation of phantom work (see Fig. 8).
Figure 8. Density and usability of a system These findings, particularly the destructive and undesirable haracteristics of phantom work cycles, are consistent with omplex interactions observed in intelligent environments [36]. ‘rom Zamudio’s observations, and the findings of our work we surmise that, in business processes, Design-Build-Tes ire necessary for new produc development, and in the cycles case of ntelligent environments cyclic and coupled interactions can rovide specific services to the final user. Extrapola ting the esults of Zamudio’s earlier work suggests that in both domains here is a threshold poin yerformance decreases drama’ © from which the system tically: in the case of in mvironments the system wi 1 self-lock, and hence telligent become musable; in business process redundant innovation cycles can ead to the creation of phantom work (see Fig. 8).
Related topics:
Business ModelingBusiness Process ManagementAgent Based SimulationSystem Dynamics ModelingBusiness Process ModelingDigital Business ModelsBusiness Process RedesigningReconfigurable architecturesStrategy (Business)Dynamical systems and Chaos
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved