Improving the Supply Chain: Analysis

This is a series on Supply Chain Basics looking at the discipline from the Society of Operations Management perspective. Supply chain is also essential to project management as PMs are typically trained in world class contracting. For example, my Masters program had several courses involving contracting and the Defense Acquisition Workforce Improvement Act, DAWIA, certification highlights the combination of project management and supply chain. In this post, we will explore analysis issues adding some additional support as well.

Analysis

Continuous improvement is directed from the top and implemented from the bottom up. Establishing the framework, selecting processes for improvement, and setting goals is a job for top management who align strategy to the direction of the organization. After this is in place then staff at all levels company wide and also across the supply chain begin to implement. The teams first effort is to properly frame the process details and then analyze the process looking for weak performing parts. These weak performing parts are then assessed for root causes or the underlying reasons for the low performance. This post will explore some of the techniques used to assess the  process and get at root causes. 

Process Mapping and Pattern Identification:  Process mapping is the broadest used and most successful method. APICS defines a process as a diagram of production or service operations that flow through a production system. Standard symbols indicate processing, flow direction, decisions, inputs and outputs, as well as other aspects of the process.  Mapping the process aides the improvement teams with identifying and flushing out inefficiencies and ineffective process areas. Useful information out of mapping includes sequencing, duration, timing, required resources, responsible parties, financials, etc... Process maps may be drafted at first using a pencil and paper, whiteboards, and dry markers as they may be outcome of storming and forming before putting them into the computer system.  Software such as Microsoft Visio, SigmaFlow, iGrafx, PowerPoint, Excel, Word, etc...

One way to conduct process mapping is to brainstorm using PostIt® Notes and a whiteboard. On the PostIt® Note the flowchart process symbols are drawn with accompanying notes and process descriptors. The Postit® Note is then stuck to a whiteboard and dry marker lines connect the PostIt® Notes. The diagram can be quickly reordered and updated as the process is discerned.  Colored connecting lines can indicate flow importance, direction, or internal versus external among other things.

After determining the process with high detail and clarity, the next effort is to begin to look for behavioral patterns  and variability in the conduct of the process.  Patterns and variability occur over time requiring the collection of data. If operating under Six Sigma the process error rates become more important and have to be tracked using control charts and defect measurements.

Control Charts: APICS defines control charts as a graphic comparison of process performance data with predetermined control limits. Control charts are associated with Six Sigma, 6σ, and Statistical Process Control, SPC.  The chart typical has performance data over time with normalized performance occurring around a centerline, CL. Variation is shown above and below the center line and is considered out-of-limits when it occurs above or below the Upper Control Limit, UCL, or the Lower Control Limit, LCL. Overall, control charts are useful for monitoring and managing a process as well as correcting out-of-limit conditions.  Figure 1 shows long standing interpretation standards for control charts.

Figure 1: Control Chart Interpretation and Evidence for Investigation
Source: Hansen, B. (1964). Quality Control Theory and Applications.  Pearson Education: NJ. p. 65.

Defect Measurements: The emphasis is on the customer's definition of a defect. This requires visibility of the customer interactions and reactions to the product or service. Defect measurement is straight forward accounting of the number of defects once a defect is described or defined.  The defect can be qualitative or quantitative. Qualitative defects could associate weight perceptions, color, or other visceral and emotional aspects of the design. Quantitative defect measurements are typically be associated with spatial and volumetric measurements but could include activities like return rates, scrap rates, or even rate of sales changes.

Root Cause Assessment

After the process is identified and the patterns are discerned, the problems or at least symptoms of poor performance  are usually observed based on the goals and objectives. The identified poor performance areas require further investigation. Discerning the causes of poor performance usually rests on root cause analysis which aides in uncovering original causes and not just the symptoms. Root cause analysis is the underlying or original source of an unacceptable rate of defects or performance. Attacking a secondary problem or a symptom will do little to correct the problem and reduce or eliminate customer complaints. Three tools used in root cause analysis are Pareto Diagrams , Cause-and-Effect Diagrams, and the 5W's.

Pareto Diagrams: This is a concept designed by the Italian Economist, Vilfredo Pareto, that states a small percentage of a group accounts for the greatest impact and value.  Related to the Pareto principle is the 80-20 Rule which states 80 percent of the impact or effects come from 20 percent of the causes.  The diagram aides in isolating possible sources of the greatest problems and is percentage occurrence against attribute.  This process can be achieved using spreadsheets. 

Cause-and-Effect Diagrams: Total Quality Management, TQM, is most associated with the diagram despite the many names such as the Ishikawa or fishbone diagram. But whatever the name the goal is to look at the basic inputs into an operation that is behind a stated problem.  The fundamental inputs are generally money, manpower, methods & measurements, machinery, materials, and sustainability issues. 

Figure 2: The Fishbone Diagram Template
Source: Concept Draw resourced 04Nov2013 from
http://www.conceptdraw.com/solution-park/business-fishbone-diagram

5Ws: Why, Why, Why, Why, and Why? The basic theory is to keeping asking 'why' until the root cause is expose. Of course, one can ask why to the point of the singularity or the origin of the Universe but the reasonable person would stop significantly sooner. Under the theory the problem is only five whys away.  

These basic approaches and methods to mapping processes, discerning process patterns, and getting at the root causes requires little overhead. Six Sigma can be time consuming with a considerable amount of overhead. Most companies are conscious of overhead and time. Applying the Pareto principle to the analysis practice, 80% of the impact can be derived from 20% effort. Thus, on-the-fly practices of brainstorming processes, using spreadsheets to collect defects, and asking why can yield effective results without the labored involvement of Six Sigma. 

Comment: Detailing organizational processes in a dynamic business environment can lead to processes that change as fast as they are mapped. In these cases, a better solution is to look at the patterns and map management of the patterns instead of a process. Then apply process to the patterns. This is and has been the idea behind the use of complex adaptive systems. The organization self organizes as emergent conditions drive patterns and information exchanges. This becomes essential in operations like crisis management where anyone of numerous contingent events can emerge. Business operations that are highly sensitive to market conditions such as LTL and airline carriers must be adapt quickly. The processes can change rapidly as vehicles breakdown, weather, and market demands shift in various combinations as well. 

Once the patterns and processes are known and under control, automating the process and tracking elemental performance is the next big step. There are numerous software applications that layer process and infrastructure in ways for decision making. For example, CAD has a version for urban infrastructure design that has lay downs for all the electrical, sewage, water, streets, etc... of the city. This has a layer for 911 data as well and can track movement of police, fire, and ambulance vehicles throughout the city. Traffic data is also fed into the system in order to intelligently control traffic flow and clear the path of enroute emergency vehicles. Likewise, supply chain processes can be mapped end-to-end increasing real time visibility. As materials and information move through the process, alerts and triggers can aide in the management of the supply chain. Thus, the analysis piece can lend well to a much larger effort. 

Reference:

(2011). APICS Certified Supply Chain Professional Learning System. (2011 ed.). Version 2.2.