OM Week 2

What was Dell’s solution to having no single source of truth?

Create two data foundations

1. Harmonize: consolidate sources

2. Visualize: single dashboard

Why can’t all companies require customers to pay upfront?

1. B2B Customers expect credit: Hospitals, governments, corporations demand Net 30/60/90

2. High Value Purchases: Boeing can’t ask airlines for $200M upfront

3. Competitive Pressure: If competitors offer financing, you must too

4. Industry Norms: Construction uses progress payments; restaurants buy ingredients before customers arrive

5. Trust and Brand Power: Unknown brands can’t demand upfront payment (Ex: A startup selling laptops - no chance)

Why did Dell’s Upfront Payment Model work?

• Individual consumers (not B2B)

• Affordable price points ($1-3K)

• Strong brand trust

• Competitors didn’t offer financing

Dell PC Assembly Process

What is the bottleneck within Dell’s PC assembly process?

Assembly

Bottleneck

The step that limits process capacity and has the lowest output rate (longest cycle time)

Why is it important to identify the bottleneck of a process?

• Bottleneck limits the entire system

• Work piles up before it

• Improving non-bottlenecks won’t help improves the processes effciency

What is the cycle time of the system?

5 minutes

Station CT Formula

Processing Time/# Resources

System CT Formula

Bottleneck’s CT (the max)

What is Cycle Time?

How often does a completed unit come off the production line?

• The “Rhythm” of production

What is Throughput Time?

Total time for one unit through the entire process

• How much time total does one unit take to be built?

• Also called Flow time, Lead time

• = Sum of all processing times

Rush Order Flow Time

Minimum throughput time (no waiting)

• In an empty system, your unit takes the combined total time of each step within the system

Cycle Time and Throughput Time are independent of one another (True/False)

True

Cycle Time depends on the bottleneck only while Throughput Time depends on the number of steps (and total time of each step)

• When you add more steps, CT stays the same (if the bottleneck is unchanged), however, the TT will increase

Output Rate Formula

Output Rate = 1/Cycle Time = 1/Bottleneck Time

Output is entirely set up by the bottleneck (True/False)

True

• Speeding up other steps WON’T help

Process Capacity

Maximum output the process can produce

Process Capacity formula

Capacity = 1/Bottleneck CT

Actual Output (Flow Rate) Formula

Actual Output (Flow Rate) =. min(Demand Rate, Capacity)

If a system is Supply-Constrained:

• What can we say about the relationship between demand and capacity

• What can we say about the flow rate?

• What is the limit of the system?

• Demand > Capacity

• Flow Rate = Capacity

• The Process is the limit

If a system is Demand-Constrained:

• What can we say about the relationship between demand and capacity

• What can we say about the flow rate?

• What is the limit of the system?

• Demand > Capacity

• Flow Rate = Demand

• The Market is the limit

Utilization Formula

Utilization = Flow Rate/Capacity

Utilization is always <= 100% (True/False)

True

What is true about a Demand-constrained process

It has idle time

Time for Q Units Formula

Time = TT + CT x (Q-1)

Cost of Direct Labor Formula

CODL = Total Wages Per Hour/Flow Rate

What is true about Buffers when there is uncertainty in the bottleneck?

It will always be useful

Rush Order Flow Time assumes No Buffer (True/False)

True

Step Capacity

What ONE step can handle

System Capacity

What the WHOLE process can handle

System Capacity Formula

System Capacity = min(all Step Capacities)

Step Capacity Formula

Step Capacity = 1/Activity Time

• Each step has its own capacity

Flow Rate is limited by Step or System Capacity?

System

• Improving a bottleneck step does NOT increase system capacity

• We assume no supply constraints

Utilization tells you…

Where your money is being wasted (idle resources) and where you’re maxed out (bottleneck)

Utilization

The fraction of capacity that is actually being used (How busy is the resource/)

Utilization reveals:

• Waste: Low utilization = paying for unused capacity

• Bottleneck: Highest utilization = the constraint

• Imbalance: Big gaps in utilization = poorly designed process

Utilization Formula (2 ways)

1. Utilization of a Step = Flow Rate (System)/Capacity (Step)

• Numerator = system output rate

• Denominator = step’s max rate

• Always works

1. Activity Time of Step/Activity Time of Bottleneck

• Numerator = time at this step

• Denominator = time at bottleneck

• Restrictions apply

• Only works when the system is capacity constrained and there is a single worker at each step (with parallel workers you must use Formula 1)

Step-by-step method for calculating utilization

1. Find the Flow Rate for the entire process: min(Demand, Capacity, Supply)

2. For each step, calculate: Utilization = Flow Rate/Step Capacity

3. The same Flow Rate goes through every activity (what goes in must come out)

Why can utilization never exceed 100%?

1. Bottleneck = step with the lowest capacity

2. System Capacity = Bottleneck Capacity

3. Flow Rate <= System Capacity

4. Therefore Flow Rate <= Bottleneck Capacity <= Any Step Capacity

So when you calculate Utilization = Flow Rate/Step Capacity

• The numerator (Flow Rate) is always <= the denominator (Step Capacity)

• Therefore, Utilization is always <= 100%

What happens to the bottleneck when the system is capacity constrained?

Bottleneck Utilization = 100%

• Working nonstop and still can’t keep up

• Limits the system

• Improving the bottleneck increases capacity

• No slack for variability

What happens to the bottleneck when the system is demand constrained?

Bottleneck Utilization < 100%

• Not at limit BUT still the highest

What is true of steps that are not the bottleneck (don’t have the highest utilization)?

• These steps have excess capacity

• Improving these steps won’t help throughput

• These steps have slack to absorb variability

What is the Cost of Direct Labor (CODL)?

The labor cost incurred to produce one flow unit

Why do managers obsess over CODL?

• Pricing: CODL sets the floor for profitable pricing

• Efficiency: Lower CODL = more efficient process

• Comparison: Compare different process designs objectively

• Decisions: Should we add workers? Combine tasks? Automate?

Every process change affects CODL (True/False)

True

Cost of Direct Labor (CODL) Formula

CODL = Total Wages per Hour/Flow Rate per Hour

• Units: dollars per unit

Step-by-step method for calculating CODL

1. Count all workers in the process (every station)

2. Calculate Total Wages/hr = Sum(each worker’s hourly rate)

3. Find Flow Rate = min(Demand, Capacity, Supply)

4. Divide: CODL = Total Wages/Flow Rate

Include ALL workers who touch the process → even if they’re not busy 100% of the time

A higher flow rate leads to a [BL|ANK] CODL

1. Higher

2. Lower

Lower

• You spread the fixed labor cost over more units

CODL goes DOWN when you (3 reasons)

• Increase flow rate

• Reduce wages

• Improve bottleneck

CODL goes UP when you (3 reasons)

• Add workers w/o more output

• Demand drops

• Overstaffing

Adding workers when demand constrained [BLANK 1] CODL because flow rate doesn’t change, you pay [BLANK 2] for the same output

1. Increases/Decreases

2. More/Less

Cost of Direct Labor (CODL) Daily Perspective Formula

CODL = Total Daily Labor Cost/Total daily Output

How does capacity change when you have multiple workers doing the same task?

It depends on the configuration

How does capacity change when you have multiple workers doing the same task?

• “OR” configuration

Each flow unit goes to ONE worker

• Workers work in parallel

• Capacity = Sum of capacities

How does capacity change when you have multiple workers doing the same task?

• “AND” configuration

Each flow unit needs ALL workers

• Workers work together

• Capacity = Min of capacities

“OR” Configuration Formula

Capacity = m/Activity Time

• m = number of parallel workers

• Adding parallel workers multiplies capacity

When you improve the bottleneck, the bottleneck often shifts to another activity (True/False)

True

• You’ve fixed one constraint, but now another becomes the limiter

If the system is Demand Constrained

• Adding workers increases cost without increasing output

• Focus on reducing costs or finding more customers

• Don’t expand capacity

If the system is Capacity Constrained

• Adding workers at bottleneck increases output

• Focus on expanding capacity at the bottleneck

• Watch for shifting bottlenecks

Flow Time

The total time a flow unit spends in the process from entry to exit

Flow time includes:

• Time being processed (activity time)

• Time waiting between activites

Two Versions of Flow Time

1. Theoretical flow time (empty system): Just activity times, no waiting

2. Actual flow time (real system): Activity times PLUS waiting times

Theoretical Flow Time is also known as

1. Rush Order Flow Time

• ROFT = Theoretical Flow Time = Sum of Activity Times

2. Theoretical Flow Time

3. Minimum Flow Time

4. Processing Time

Theoretical Flow Time

Time for ONE flow unit through entire process when system is empty

Step-by-step ROFT calculation

1. List all activities in sequence

2. Sum all activity times

ROFT [BLANK] change when you add parallel resources

1. Does

2. Does Not

Adding a second worker at Activity B doesn’t make B faster for any single unit (it just means B can handle more units per hour)

What is the key distinction between Cycle Time vs Flow Time

Cycle Time is the time between successive outputs while Flow Time is the time one units spends in the system

• Cycle Time tells you the rhythm of the output while Flow Time tells you the customer experience

Throughput Time

The total time needed to complete all N flow units that arrive during a period

Once the process reaches a steady state, units exit at the

Flow Rate

Throughput Time Formula

Throughput Time (Time to Process N Units) = N/Flow Rate

Step-by-step Overtime Calculation Method

1. Count total units: How many flow units arrive during the shift?

• N = Arrival Rate x Operating Hours

2. Calculate time needed: How long to process all N units?

• Time Needed = N/Flow Rate

3. Calculate overtime: Compare to available hours

• Overtime = max(0, Time Needed - Operating Hours)

If Time Needed <= Operating Hours, then Overtime =

0

When you change a process, what signals a Good/Bad Change:

• Flow Rate

• CODL (cost per unit)

• Overtime

• Utilizations

• Meets demand?

Flow Rate

• Good: Increased

• Bad: Decreased or unchanged

CODL (cost per unit)

• Good: Decreased

• Bad: Increased

Overtime

• Good: Decreased

• Bad: Increased

Utilizations

• Good: More balanced

• Bad: More unbalanced

Meets Demand

• Good: Yes

• Bad: No

A good process improvement should (3 points)

• Increase throughput

• Reduce cost per unit

• BOTH

Before recommending a change

1. Calculate Base Case (current state)

• Bottleneck, capacity, flow rate

• Utilizations at each step

• CODL, overtime, total daily cost

2. Proposed Change

• New bottleneck (did it shift?)

• New capacity, flow rate

• New utilizations, CODL, overtime

3. Compare

• Did flow rate increase?

• Did CODL decrease?

• Did overtime decrease?

• Does the system now meet demand?

The 7 Step Process Analysis Method

1. Define the flow unit: What moves through the process?

2. Measure activity times: How long does each step take?

3. Calculate capacity: Capacity = m/Activity Time

4. Find the bottleneck: The step with the lowest capcity

5. Determine flow rate: Flow Rate = min(Demand, Capacity, Supply)

6. Calculate utilization: Utilization = Flow Rate/Capacity

7. Calculate CODL = Total Wages/Flow Rate