2 Process Analysis and Improvement

What a process does

A process transforms inputs—such as materials, information, or customer requests—into outputs through connected activities. Process analysis examines how work actually moves through those activities, how well the process performs, and where changes could improve results.

Process types and operating choices

Process types differ mainly by volume, or how much they produce, and variety, or how much outputs differ. As volume rises and variety falls, processes generally become more standardized and equipment more specialized.

• Project: Very low volume and highly customized work; activities may vary by output. Constructing a bridge is an example.

• Job shop: Low volume and high variety, with work following different routes. A repair shop handling different jobs is an example.

• Batch: Moderate volume, with groups of similar items processed together. A bakery making several trays of one product is an example.

• Repetitive or line: High volume, with standardized work following a recurring sequence, as in an assembly line.

• Continuous: Very high volume and highly standardized, uninterrupted flow, as in oil refining or water treatment.

These categories form a continuum, and real operations may combine types. A business should select a process that fits its needed mix of cost, flexibility, and output variety.

Mapping the work

A process map makes the steps and connections in work visible. It helps people who perform or use a process check how work really happens, rather than relying only on how a procedure says it should happen. Mapping can reveal delays, repeated work, unnecessary steps, errors, and handoffs.

A practical mapping sequence is:

1. Set boundaries: Define the starting event, endpoint, and scope.

2. Identify participants and inputs: Include people who do the work and those who provide or receive inputs and outputs.

3. Map the current process: List the steps in order, including decisions, rework loops, and waiting.

4. Verify the map: Walk through the process with the people involved and correct gaps.

5. Mark evidence: Add step times, queues, errors, or other measures where useful.

A high-level SIPOC map covers suppliers, inputs, process, outputs, and customers, and can help define the process before drawing a detailed flowchart. A value stream map extends the view across connected steps, showing material and information flow as well as opportunities to reduce waste and waiting.

Measuring flow and capacity

Flow analysis examines how work moves through a mapped process. Useful measures distinguish active work from idle time and describe output, work inside the process, and the limits on production:

• Processing time is the time spent actively working on a unit.

• Waiting time is the time a unit spends idle, often in a queue.

• Throughput is the number of completed units per unit of time.

• Work in process (WIP) is the number of units currently inside the process.

• Capacity is the maximum output rate a step or process can sustain under stated conditions.

For a simple sequential process, estimate a step’s capacity by dividing its available work time by its time per unit:

In practice, downtime, staffing, product mix, and variability can affect actual capacity, so estimates should be checked against observations.

Finding and managing bottlenecks

The step with the lowest effective capacity is the process’s bottleneck. Assuming demand is sufficient, it limits the maximum throughput of the whole system.

For example, a clinic can register $12$ patients per hour, examine $8$ per hour, and check out $15$ per hour. Examination is the bottleneck, so under these assumptions the whole process cannot sustainably complete more than about $8$ patients per hour. Adding checkout capacity alone will not raise total throughput, and patients may instead wait before examination.

Bottlenecks can shift when conditions or process capacity change. Improving a step that is not the bottleneck may reduce its local workload without increasing total output. The Theory of Constraints emphasizes identifying and managing the constraint while improving the system as a whole.

Testing and sustaining improvements

Begin with a clear performance goal, such as shorter lead time, higher throughput, fewer errors, or more reliable service. Use process maps and measured data to identify causes instead of assuming that the most visible delay is the root cause. Possible changes include removing unnecessary steps, reducing handoffs or setup time, balancing work, preventing errors, or improving information flow.

Plan–Do–Check–Act (PDCA) provides a disciplined way to test changes:

1. Plan: Define the problem, establish a baseline, and choose a change to test.

2. Do: Try the change, preferably on a small scale.

3. Check: Compare results with the baseline and check for unintended effects.

4. Act: Adopt and standardize a successful change, or revise it and test again.

Continue measuring after implementation. A change that increases speed but harms quality, safety, or customer experience is not a sound improvement. After relieving one constraint, reassess the process because another step may become the bottleneck.