Industrial engineering focuses on improving systems that combine people, machines, materials, information, and time. Six Sigma is a structured approach for reducing defects and process variation. This cheat sheet helps students connect engineering design decisions to measurable performance, cost, quality, and efficiency.
It is useful for analyzing factories, hospitals, service systems, supply chains, and school operations.
Key Facts
- Takt time = available production time / customer demand, and it shows the pace needed to meet customer demand.
- Cycle time is the time required to complete one unit or task, and a process can meet demand when cycle time is less than or equal to takt time.
- Capacity = available time / cycle time, and the station with the lowest capacity is usually the bottleneck.
- Utilization = actual output / design capacity x 100%, and very high utilization can increase waiting time and delays.
- Process yield = good units / total units x 100%, and rolled throughput yield = yield 1 x yield 2 x yield 3 for a multi-step process.
- DPMO = defects / (units x opportunities per unit) x 1,000,000, and it measures defects per million opportunities.
- Cp = (USL - LSL) / (6 x sigma), and Cpk = minimum of (USL - mean) / (3 x sigma) and (mean - LSL) / (3 x sigma).
- DMAIC stands for Define, Measure, Analyze, Improve, and Control, which is the standard Six Sigma improvement cycle.
Vocabulary
- Industrial engineering
- A branch of engineering that designs and improves systems involving people, materials, equipment, information, energy, and time.
- Six Sigma
- A quality improvement method that uses data to reduce variation and defects in a process.
- Bottleneck
- The process step that limits the total output because it has the lowest capacity or longest effective cycle time.
- Takt time
- The maximum time allowed to produce one unit in order to match customer demand.
- Defect
- Any result, product, or service outcome that does not meet a required specification or customer expectation.
- Control chart
- A graph used to track process data over time and identify whether variation is stable or unusual.
Common Mistakes to Avoid
- Confusing takt time with cycle time is wrong because takt time comes from customer demand, while cycle time comes from the process itself.
- Ignoring the bottleneck is wrong because improving a non-bottleneck step may not increase total system output.
- Using average output only is wrong because variation, downtime, and defects can make a process unreliable even when the average looks acceptable.
- Calculating DPMO without including opportunities per unit is wrong because one unit can have more than one possible defect location.
- Assuming a process is capable just because most items pass inspection is wrong because capability compares process spread and centering to specification limits.
Practice Questions
- 1 A workstation has 420 minutes available per day and customer demand is 140 units per day. What is the takt time in minutes per unit?
- 2 A process has three stations with cycle times of 3 minutes, 5 minutes, and 4 minutes per unit. Which station is the bottleneck, and what is the process capacity in units per hour?
- 3 A factory makes 2,000 units, each with 4 defect opportunities. Inspectors find 32 defects. What is the DPMO?
- 4 A team wants to reduce delays in a hospital check-in process. Explain why measuring variation and identifying the bottleneck should come before choosing a solution.
Understanding Industrial Engineering & Six Sigma
A process is more than a list of steps. It has inputs, work, decisions, movement, waiting, inspection, and handoffs. A useful process map shows where each item goes and who or what performs the work.
Students should separate value creating work from necessary support work and pure waste. Drilling a required hole adds value to a part. Searching for a tool or moving a box across a building usually does not.
In an office, entering the same customer data into two systems is a common source of waste. Mapping the real process matters because the written procedure often differs from what people actually do.
Flow becomes difficult when work arrives unevenly or when steps take different amounts of time. A slow station can create a growing line before it, while later stations may sit idle. Adding workers at an already fast step will not fix that problem.
Improvement should focus first on the constraint that limits the whole system. Variation makes this harder. Even when the average work time seems acceptable, a few long tasks can cause queues and missed deadlines.
High utilization may look efficient on a report, yet a nearly full system has little room to absorb a machine failure, a late delivery, or an unusually complex order. Some spare capacity is often a practical way to protect reliable service.
Quality data needs careful interpretation. A defect is a failure to meet a requirement, but one unit can have more than one possible defect location. That is why defect measures must clearly define the unit, the opportunity, and the inspection method.
A paint chip, a missing label, and an incorrect dimension may count differently depending on the process. Capability measures compare the natural spread of results with the allowed specification range. A process can have a narrow spread but still produce bad parts if its average is shifted too close to one limit.
Students should remember that customer specifications come from product needs. Control limits come from observed process behavior. They answer different problems and should not be treated as interchangeable.
DMAIC provides discipline when a team wants to improve a process. In the Define stage, the team states the problem in measurable terms and identifies who is affected. Measure requires trustworthy data, since poor measurement can create a false problem.
Analyze looks for causes using evidence rather than guesses. For example, defects may rise only on one shift, with one material batch, or after a certain machine setting changes. Improve tests a change on a limited scale before full adoption.
Control keeps the gain through standard work, training, checks, and clear response plans. Control charts are especially useful because they show data over time. Random small movement is normal common cause variation.
A sudden jump, trend, or repeated pattern can signal a special cause worth investigating. This approach applies to assembly lines, cafeteria queues, lab experiments, package delivery, and group projects where late handoffs slow everyone down.