Commentary: This a further discussion on the Operations Management series regarding complex adaptive systems. Once the nodal conversion is made, the next level is to manage the communications and organization of the nodes. The why, who, where, what, and when will be addressed regarding the management of the network.
Complex Management Adaptive Management
Under traditional Six Sigma and lean events, an inefficiency is identified then efforts are under taken to correct the inefficiency and then the event is complete as the team moves on. The identified inefficiency is found using the DMAIC process and to some extent preference. Many of these inefficiencies hae arisen out of poor implementation of new projects and organizational design due to ever changing leadership vision and the rush to results. Other ineffiencies arise from regulations and environmental conditions that force fit or drive awkward processes and operations. In order to prevent introducing inefficiencies, Six Sigma has the DMADV process. The DMAIC and DMADV methods are quality processes that have been extended to business processes other than manufacturing processes. Six Sigma and lean projects are continuous; require costly labor, certifications, and time; and are limited to identified issues that can be subjective despite the intensive analytical efforts involved. Both the DMADV and DMAIC processes are cause and effect based and outside the normal operations. What if the organization was adaptive to new projects and environmental conditions by design in the first place? This is possible with complex adaptive systems.
If designed correctly complex adaptive systems can respond to changes in near real time. One key metric is organizational latency which is the time to correct a problem from the instant of identification to the instant of recieving the full benefit. Minimizing organizational latency saves cost and strengthens time-to-market / impact efforts. Organizational latency under Six Sigma and lean is very high because the process is external to the operation. Teams outside the process look at the situation in order to determine the corrective action then apply that to the operation at some point in the future. That is Six Sigma and lean are applied to the operation rather than a part of the operation. The identification of a problem, the measurements, and analysis are required because the team is usually not familiar with the situation and is seeking confirmation and identification of the suspected issue. Complex adaptive systems are designed in a way that adaptive nodal networks adjust to changes as they emerge as opposed to traditional organizational approaches that have huge latencies.
Nodes are assigned a unique set of related activity-processes and are the single source for the organization for those activity-processes. There is a irreducible state in order to provide the activity-services supported by the node. Contraction beyond that state either causes the node to be outsourced or otherwise non-functional. Expansion of a node adds discrete levels of staffing and/or a parallel node performing the same function increasing capacity of the activity-processes serviced. Entirely new nodes change the capability of the organization. Design of a new capability requires a DMADV like process that also includes a make, lease, or buy decision for the set of processes and activities of a node. Some capabilities can be outsourced and integrated into the network through control points. Nonetheless, adding capability could be fully encapsulated in a single node or integrated across several nodes based on the nature of the capability. More than likely the capability can be fully encapsulated in one or more additional nodes that draw and/or provide information and materiel support to other existing nodes. Overall, once a node is functional and in the network, decisions regarding information sourcing and communications are generally made locally within the node. They would know which other nodes provide specific information and material support they need then can connect to those nodes in support of the work effort in sort of a interesting variation of datamining. Reporting systems can collect products and performance data from a node's output.
Management of control points and the network requires a virtual model of the business communications and nodes. Management of the operations would require supervisor nodes that are virtual and monitor the network activity in order to maintain catalogues and performance data. At this level of management decisions are made about nodes and the communication channels. Financial management can occur along the way rather than top down as each node is an economic center buying inputs and selling outputs to other nodes. Full financial reporting can occur at the node level then rolled up to the whole. Resources can be assigned economically to the node. Nodal balance sheets could reflect retained earnings to be applied towards node development.
Overall, the idea is to create a operational system that is self-organizing (using supervised intelligent systems) and correctly managed at the appropriate levels. When the system can realign and readily change their information and materiel sources at the right level enormous efficiencies become possible. Discovered shortages and deficiencies as well as surpluses become immediately apparent and addressed in terms of adjustments to capabilities and capacities. Essentially, much of the define, measure, analyze, improve/design, verify, and control process becomes embedded or innate to the system in various forms as opposed to something that happenes to the system. When this happens the organizational latency shrinks, responsiveness or adaptibility increases, and sustainability becomes a way of operating.
I will hold off on posting further posts to this series for now in order to focus on other topics. I may add new posts to this series as people comment or email concerns to me.
No comments:
Post a Comment