Do you know how you fail?
The manufacturing industry has been evolving out of standard productivity models to higher intelligent systems based on this human experience of "seeing it early", or "knowing the fail".
Prevention is better than cure, goes the old adage and all of us experience in our lives that if an event (especially one which creates damage) can be seen or perceived early, the job of managing it is half done. The time, energy and investment required to fight a "bottleneck" can easily be avoided if we can prevent that same event from happening or if we know the point of occurrence, and its extent (breadth and depth).
The manufacturing industry has been evolving out of standard productivity models to higher intelligent systems based on this human experience of "seeing it early", or "knowing the fail". Ever since the industry struggled with putting in all relevant measures or getting equipped with ammunitions to fight "bottlenecks" when they set in, there has been no end to failures. Failures happened, got repeated and varied with time, and in each such event the situations were attributed to machine, schedules, material, demand forecast and also the type and nature of human interactions. The net result - this vicious circle does not see any light and a huge investment on managing failures goes to waste. Many times the measure proves unproductive on account of the lack of knowledge of predictable outcomes of such failures.
The tricks of technology deployed on the manufacturing value chain has been doing quite a bit in making life relatively easier for engineers and production planners, with optimization of human interface in many areas. However the systems and tools in place have not been able to bring a permanent solution to the problem of "preventing failures". Starting from the non-availability of spares in maintenance activities to the non-utilization of production lines, inventory pile ups on the floor, and missing of shipment deadlines, the resilience of the supply chain still does not seem enough.
Probably predictive analytics seems to be the sun in the horizon to make the industry see a new dawn. It brings with it the promise of its self-grown knowledge base gathered over time, or by inputs of various system and human interfaces present in the current process which makes it raise the red flag much before the bottleneck rises to the surface. This is being made possible by its ability to have a real time integration with all the parts and elements of a supply chain - e.g. for an order placed or being put in system for fulfillment, the past history of the relevant applicability parameters in terms of demand forecast alignment, sourcing lead times, manufacturing lead times, shipping dependencies etc are all verified so as to ensure a no-failure path ahead.
The opportunity seems appropriate with the industry seeing some hope to get itself free from the clutches of never ending failures. However the success of this framework would depend on the rate and scale of adaptability in the processes, and a positive mindset of all the stakeholders to contribute their own part.
The following are broad areas in which the manufacturing industry is expected to see the positive impacts or outcomes of this analytics framework:
Quality procedures and methodologies - A more varied mix and volume of data integrated from various elements is supposed to provide some definite insights on "parameters of measure" and lead to desired improvement.
Supply chain transparency - The demand forecast versus resource availability visibility at every stage of processing would ensure the necessary intelligence to plan for the resources - man, material, machine, money better and would also help "predict" a red flag of one step forward which is directly or indirectly dependent on failure of the current step.
Reduce or eliminate machine idle time - With more visibility of resources and desired outcomes (unified database) the production scheduling will become more intelligent generating higher or accurate utilizations.
Preventive Maintenance - With resources and utilizations in sync with the plan and performance of various elements - machines, systems, tools being monitored, breakdowns can be predicted based on historical information on "patterns of failures (knowledge base)"
The promise is huge. What's needed now is effective implementation and collaboration for developing a self-sustainable model in the manufacturing value chain which can say - "I know you are going to fail this way there".... because "... what you have been doing till now has its seed here ...," and help in delivering desired outcomes seamlessly.