BE1401 Peer Tutoring

Final Quiz Summary Slides

Introduction to Processes
Set of trade offs: Three system inhibitor (that cause waste of resource): Two types of cost:
-Cost -Waste • Input (raw materials)
-Variety -Variability (Demand variability vs Supply variability)
-Quality -Inflexibility • Resource (those that
-Time transform input to output)

Pareto Domination (Inferior to one more competitors in ALL dimensions): Little's Law:
A B C
I=RxT
Speed 6 10 1 (Inventory = Flow rate x Flow time)
Cost $3 $5 $10
WIP = Throughput x lead time

WIP- Average flow units

• C is pareto dominated by A and B
Throughput- Average depart rate

• A and B are at efficient frontier Lead time-Average time

(not pareto dominated) unit spends in system

• Inefficiency - gap between firm and

efficient frontier
Process Improvements
Process Improvements
Process Interruptions

Multi-Station Process (E.g. Xootr)

• Important to analyze by flow units (Tabulate capacities by station)
• If 1 unit of product consist of 1 part of A and 2 parts of B, then A+2B is 1 flow unit
• Flow rate (of 1 unit of product) = min{demand, process capacity} Note:
• Remember to factor in capacity of assembly

Multi-Product Process (E.g. Chicken + Tomato + Onion Soups)

• Analyze by total batch size per production cycle
• Sum of all setup time in production cycle
• Flow rate = min{demand, process capacity}
• When calculating inventory for 1 of the product, all values used should be with respect to that product
• After calculating the total batch size, take weighted average to find individual batch size
Queue Systems & Analysis
Queue growth rate = Demand – Capacity To reduce demand:
• Peak load pricing-charge more during peak
Length of queue at time T = T*(Demand-Capacity) • Off peak disc-offer discount during low demand

Time to serve Qth customer = Q/capacity To adjust capacity:

• Pre-process-shift work from peak to off peak if
Time to serve person arriving at time T= T*((Demand/Capacity)-1) possible

Average time to serve a unit = (1/2)*T*((Demand/Capacity)-1)

Single server model: Multiple server model:

For exponential distribution, CV = 1 Unstable queue: demand > capacity [p<a]

Iq = (1/a)(Tq) [Inventory in queue]

Ip = (1/a)(p) [Inventory in service]
Iq + Ip = Total inventory in the entire process Stable queue: demand < capacity [p>a]

Effects of pooling: reduces a and Tq

Project Management
Critical path: Path that takes longest time Three estimates:
• List out all possible paths and calculate their time • Optimistic time (a)
• Activities on critical path have 0 slack • Most likely time (m) Mean:
• Pessimistic time (b)
Slack: Maximum time an activity can be delayed
• LS - ES te is the expected time to Variance:
• LF - EF complete that activity
Let T = Project Duration
-Assume T is normally distributed with mean TE

EF= ES + t -Assumes activity times are independent

LF= LS + t
ES: max EF of all preceding activities
LF: min LS of all successor activities Analyzing probabililty of project completion:
Linear Programming
Feasible solutions: Solutions that satisfies all constraint

Infeasible solutions: Solutions that violates at least 1 constraint

Optimal solution: Feasible solution that maximises objective function

E.g.
Possible types of questions:

Objective function 6X+5Y​

X+Y ≤ 5​
Constraints 3X+2Y ≤ 12​
X ≥ 0, Y ≥ 0​

Feasible solutions include: (2,2), (0,5) [Satisfies all constraint] Find the values of S, T and U

Non-Feasible solutions include: (5,0), (7,0) [Violates at least one or all constraints]
• Understanding scheduling problem
However, only (2,3) is the optimal solution [Maximizes objective function]
Forecasting
Types of error calculation:ty Types of forecasting:ty

Naïve: [Significantly influenced by statistical noises]

Moving average: [different weights may be used]

Exponential smoothing:

Double exponential smoothing:

1. Forecast for demand:

2. Forecast for trend

Intro to Inventory management
Reasons for holding inventory: Types of inventory holding cost:
-Flow time -Pricing -Opportunity cost -Obsolescence cost
-Seasonality -Batch -Inventory storage cost -Spoilage and shrinkage cost
-Buffer -Stochastic Demand
Days of supply
-Average amount of time it takes for a unit to flow through the system

Inventory turns
-Number of times the average inventory flow through a process in designated intervals of time
Days of supply (1st formula) 𝐼
𝑇 = 𝑓𝑙𝑜𝑤 𝑡𝑖𝑚𝑒 =
𝑅
Days of supply (2nd formula) 1 𝐼𝑛𝑣𝑒𝑛𝑡𝑜𝑟𝑦
365 × = 365 ×
𝐴𝑛𝑛𝑢𝑎𝑙 𝑇𝑢𝑟𝑛𝑠 𝐴𝑛𝑛𝑢𝑎𝑙 𝐶𝑜𝑠𝑡 𝑜𝑓 𝐺𝑜𝑜𝑑𝑠 𝑆𝑜𝑙𝑑
Daily turns 𝑅 1
=
𝐼 𝐷𝑎𝑦𝑠 𝑜𝑓 𝑆𝑢𝑝𝑝𝑙𝑦
Annual turns 𝐴𝑛𝑛𝑢𝑎𝑙 𝐶𝑜𝑠𝑡 𝑜𝑓 𝐺𝑜𝑜𝑑𝑠 𝑆𝑜𝑙𝑑
𝐼𝑛𝑣𝑒𝑛𝑡𝑜𝑟𝑦
𝐴𝑛𝑛𝑢𝑎𝑙 𝐻𝑜𝑙𝑑𝑖𝑛𝑔 𝐶𝑜𝑠𝑡 𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒
Inventory holding cost as % of COGS =
𝐴𝑛𝑛𝑢𝑎𝑙 𝑇𝑢𝑟𝑛𝑠
Inventory management with Steady Demand (EOQ)
EOQ Model Quantity Discount

Quantity Constraint
Evaluate EOQ

Find x1 and x2, which are integer multiples of the quantity

down and up

Calculate cost (ordering + holding only) using x1 and x2

Choose x1 or x2 that gives the lower cost

Inventory management with Perishable Demand (Newsvendor Model)
Underage cost (Cu) : Cost of ordering one too few Critical ratio :
• Cu = revenue – purchase cost

Overage cost (Co) : Cost of ordering one too many where

• Co = purchase cost – salvage value

Salvage value: value that can be obtained at end of selling

season

For
Instock probability =F(Q)
Stockout probability = 1-F(Q)

Maximum profit =
Expected demand * Profit per unit sold

Expected profit =
Maximum profit – Mismatch costs

demand uncertainty, expected profit

Mismatch cost:
Gap between max and expected profit