1&2: OSCM Introduction  SCM is multidimensional

 What is OSCM strategy? What is the value of having a clear strategy?

o OSC strategy: sets policies and plans that will guide the use of the resources needed by the firm to implement
its corporate strategy. The strategy coordinates operational goals with those of the larger organization.
o The firm needs to define a clear set of priorities to help guide the implementation of a plan. A successful
strategy will anticipate change and future needs and formulate new initiative in response.
 How does strategy differ from operational effectiveness?
o Operations effectiveness: performing activities in a manner that best implements strategic priorities at
minimum cost.
o Minimize total cost of activities without compromising customers’ needs.
 Supply chain managers should not look at snapshots, but should look at the entire dynamic picture over time.
 Efficiency: doing something at the lowest possible cost
o efficiency in storage  cubed watermelons
o cost saving (purchasing, R&D, storage, continuous improving “Kaizen”  they can help)
o e.g. P&G powder  problem with transportation & forklift
 Effectiveness: Doing things that will create the most value for the costumer.
 Reduce inventory complexity: have a high percentage of common parts that compose your products (easy planning &
inventory reduction).
 Space, transportation = convenience of small packages
 IoT integration (is changing also management) / new type of services (e.g. Philips is selling lux and no more lamps…)
 Most important management issue in SCM: flexibility (especially time-based flexibility)
o Movement from cost-based to time-based flexibility
 Automation: quality, speed, 24/7 (reduce time and costs), diversification
o machines have a limit capacity  with people you can add simply other
people and produce more
 Disintegration: to be neither horizontally integrated nor vertically integrated.
 Advantages: More flexibility and agility (blurred relationship
between competitors).
 Disadvantages: Coordination, collaboration and communication
o Vertical integration: Kodak
 in the headquarter there was everything, centralize planning, clear control
 Kodak own any member in the chain
 Couldn't move fast, Apple is more flexible and can quit a relation with a supplier faster
o New type of integration (supply chain): Apple (iPhone)
 Apple is managing all the supplier  a lot of coordination
 Apple is dependent on the supplier (the same supplier has different customer  Apple, Samsung)
 Apple can find the best supplier for each part
 No big fix costs because there are no big investments
 Efficiencies are much higher
 As Apple is dependent, higher risk of copy form other producers
 Smaller player that are focusing in certain activities  Apple is just coordinating everything
 Lack of collaboration result always in lack or excess of inventory
 Outsourcing leads to Disintegration: you do not have ownership, that's the difference (in outsourcing you still have
ownership)
 Mass customization: double edge sword.
 Time is the competitive advantage (responsiveness and ability to adapt to change)
 What-If scenario analysis and contingency plans = Competitive Strategy!
 What are the challenges of OSCM:
+ More Players
+ More Products
+ Disintegrated chain (Coordinating the relationships between mutually supportive but separate organizations, ex:
outsourcing)
+ Dispersed chain (optimizing global supplier, production and distribution networks)
+ Short Life Cycles
= Increased Business Dynamics => Continuous Change => Continuous Optimization of New Decisions
 Supply chain is a strategic activity. Looking at the big picture and see how things are developing (eco system) //
network of entities  add value in the network

1
 Supply Chain Management definition: is the process of strategically managing a network of entities that are involved,
through upstream and downstream linkages, in the different processes and activities that produce value (products
and/or services) in the hands of all customers (the intermediate and ultimate consumers). Entities may include:
suppliers, carriers, manufacturing sites, distribution centres, retailers, customers and recycling agent.
o SC problem: in EU we have a not homogeneous system of transportation, and also law and regulation are
different country by country // much more cross stocking in the US than in EU // in EU space is more
expensive, high cost of warehousing
 Why companies fail:
1. Not the right KPI's/measurement systems (lack of knowledge and strategy, ex: cost accounting)
2. Culture and Organization structure (wrong organization structure)
3. Collaboration problems cannot be solved easily
4. Pascal law (pressure on one supplier, everybody will feel it… represent disintegration --> the decision that you
are taking, will impact all the chain)
5. Supplier: if you push on time  cost goes up, you stress your supply chain  quality goes down
 Key responsibilities/priorities of Supply Chain Management:
1. Value-Adding & create wealth
 operational supply chain
 meeting customer expectation and exceed them
 all of nominators are customer experiences (quality, ...)
 target pricing VS value pricing
 Value: Lower Price, Higher Quality, Better timeliness, Greater flexibility

Quality + Timeliness + Flexibility + Innovation 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒

Value = =
Price 𝑃𝑟𝑖𝑐𝑒

2. Organizational alignment (integrate activities) // interconnection

 Organizational alignment, usually horizontal, focused on corporate strategy
 Avoid twilight zone!
3. Economic efficiency
 Economic efficiency: improve production/service function of the company. Desirable relationship
between cost of capital and cost of labor.

 Type of product:
o Red box: airplanes, are very complex with a lot of supplier
 CSR: it is a priority for this box
o Fashion product / jobbing: blind buying, you don't know the demand
 it has to be fast
 supplier relation mng, customer relation mng are important
o Durable: cars, computer
 supply chain relationship is important
o Commodities: lamps, toothpaste
 capacity utilization maximization, …
 3 C’s of Supply Chain: Collaboration, Coordination, Communication
 Demand creation (vision and mission)  fulfilment (supply chain strategies, design & planning)  value chain (design
of measurement system)

2
3: Restructuring Operations (Capacity and Layout)

 Restructuring is slow and depends on your business model, it requires redesign

 Restructuring is a part of the strategy of the business.
 You need to restructure every two years
 To restructure you need to:
o understand what is your business model
o what is your dominant orientation
o pattern of diversification
o corporate attitude towards growth
 Operational competitiveness:
o Stage 1: Internally neutral — minimize negative impact of operations
o Stage 2: Externally neutral — achieve parity with competitors
o Stage 3: Internally Supportive - provide credible support to business strategy
o Stage 4: Externally supportive — create competitive advantage
 To restructure you need priorities: Rolls Royce is quality & flexibility, Seat is price
o Customer solutions (Amazon)  relationship owner  reducing customer costs, or increasing profit
o System lock-in (Intel, Microsoft)  value network architect  competition based on system economics
o Best product (Nokia, Sony)  product innovator  competition based on product economics
 Process selection & System Design => Strategic scenario planning
o Expected life time and development lead-time impacts layout choice.
o Process selection: we are looking at the lifetime cycle and on the
flexibility of the product (short life = flexible plant)
o Capacity planning: what has to be the level of automation?
o Changing the shape of the plant I can adjust the capacity! (example of
tables in a room) --> you need modelling to change things
 Layout design  product complexity VS product volumes:
o Engineer to order, configure to order, assemble to order, make to stock,
continuous (flow process)
 Product Design Timescale & Layout
o E.g. new medicines take a lot of time for design
 Productivity: outputs/inputs  input can be misleading  bad results
o Partial measure: single input
o Multi-factor measure: multiple input
o Total measure: total input
 Strategically the decision is: make or buy?
o Available capacity
o Quality consideration
o Nature of demand
o Cost
 Capacity: amount of inputs available relative to output requirements a particular time
o Design capacity (nominal capacity, not achievable) Effective capacity  Actual output
o Strategies: lead, lag (opportunity cost, I can lose share with other competitors), average, incremental VS step
by step
o Uncertainty capacity: manufacturing, customer deliveries, supplier performance, customer demand, …
 Flexibility: Capacity reserve! Capacity cushion: level of capacity in excess of the average utilization or level of capacity
on excess of the expected demand. A cushion is required to handle uncertainty in the system.
 Cushion = 1 - (Best Operating Level / Capacity Used)
 Efficiency = Actual output / Effective capacity
 Utilization = Actual output / Design capacity
 Little Law: I (inventory) = TR (throughput) * CT (cycle time)
=> any two can be managed independently. The third is naturally determined.

o if you fix the lead time you need a flexible capacity

o if you fix the capacity, you will have a variable lead time

 once you know the new process, you know to need priorities for you (costs, ....) and from that you decide what type of
supply chain

3
4: Beer Game
Manufacturer -> Distributor -> Wholesaler -> Retailer
 Capacity as a Key Issue. Synchronise all steps
 Satisfy demand at a minimum cost! Especially inventory cost.

Learned from the game: Disintegration!

 Dynamic chain, many players, disintegration
 Communication is a requirement
 Collaboration is a requirement
 Bullwhip effect is a consequence — the phenomenon in which the variability of the demand orders in a supply chain is
magnified as the orders move up to the supply chain.

Bullwhip Causes:
Bullwhip Consequences:
 Lack of information sharing
- Manufacturing cost
 Ordering in large lots - Inventory cost
 Large replenishment lead time - Transportation cost
 Rationing and shortage gaming - Level of product availability
 Price fluctuations - Labor cost for shipping and receiving
- Relationship across the supply chain
Bullwhip Avoid It:
 Align goals and incentives across functions
 Improving information accuracy (implementing collaborative forecasting and planning)
 Reduce replenishment lead time
 Reduce lot sizes
 Trust

How to change a mental model: Personal measurement, motivation, structure

 Structure is a key issue in the supply chain. CODP (Customer order decoupling point) properly positioned and sized.
Bigger Bullwhip fluctuations along the supply line (Retailer low, Wholesaler medium, Distributor, higher Factory
biggest).

Lessons learned from the beer game:

 Information loss upstream the supply chain --> efficient communication and information sharing
 Supply chain structure --> supply chain redesign: processes, task and roles (reduce lead times)
 Local optimisation --> cooperation/collaboration to achieve global optimisation
 the shorter the lead time (distribution, transportation) it will reduce the size of the inventory
 the forecast could be done in every point of the SC
 order sizes are very important --> inflated order size
 after a marketing campaign we have a drop of the demand --> it's a problem with perishable goods but also with
flowers (e.g. Heineken after a campaign continues the production, and makes a lot of stock)

---------- based on forecast (push) --------> INVENTORY <-------- based on order (pull) ------------

4
5&6: Demand Management + Forecasting

 Demand Management aim: coordinate and control all sources of demand (inside and outside organisation) to enable
efficient use of operations and timely delivery of the product. Requires a lot of coordination.
o Demand is not the demand only of the final customer, but also of demand from different sources: marketing (sales & new
product), customer service (spare parts), factory inventory (supply), manufacturing (stock)
o Strategic or tactical, the forecast of the demand is always important
 Demand management: strategic decision-making process = matching capacity and anticipate demand. Demand
forecasting is the first crucial step that must be taken, and it depends on the lifecycle where you are.
 Forecasting (Who uses it and why):
o Long-term strategic planning. Forecasting is critical in determining how much inventory to keep to meet
customer demands.
o Finance and management control (corporate planning and control, budget and cost control)
o Marketing and sales (plan and launch new products, compensate sales personnel)
o Production and operations (supplier selection, processes selection and adaptation, capacity planning, facility
layout and inventory management)
 Demand Forecast: Independent vs Dependent
o Focus on independent: the company can either influence demand (active role in managing demand, ex:
promotional campaigns, price increase/decrease, incentive to sales department --> to grow sales in December
to close the budget) or meet demand (production planning and supply chain).
 Demand Components  if company is passive, more effects:
o Trends, Seasonal factors, Cyclical elements (can be predicted)
 + average value of demand + potential trend + seasonality (Christmas, discount in January) + critical events
(difficult to predict, macroeconomic factor, wars, elections) + casual variability (hardest component,
unexplained variation) + auto-correlation
 Short term forecasts for tactical decisions such as replenishing inventory, measuring variability in demand which is
useful to set safety stock levels.
 Forecast depends on: forecast time horizon; availability of data; required precision; size of the estimated budget;
availability of qualified personnel.

1) Simple moving average (average past demand):

 Demand for a product is neither growing nor declining rapidly, not seasonal
 This method removes random fluctuations for forecasting
 Disadvantage: carry large amount of historical data
 Low n  more active // High n  more smooth

2) Weighted average (recent data is more significant than older data):

 Small n = more reaction
 Large n = smoother/flat reaction
 Disadvantage: carry large amount of historical data, more decision to take

3) Exponential smoothing - Most used demand forecast + cheapest option.

 Future demand = old forecast + error made in the forecast
 You need:
o Ft-1: most recent forecast (e.g. forecast from last week that we did for this week)
o At-1: the actual demand that courted for that forecast period
o Smoothing constant 𝜶 (controls the speed of reaction to differences between forecasts and actual demand)
o Ft = forecast for next period
 The equation states: new forecast is equal to the old forecast plus a portion of the error (the difference between
previous forecast and what actually occurred)  alpha: how much of the error I want to consider
 How to choose a value for alpha 𝛼 (range from 0 to 1):
o Low values of alpha produce smoother/flatter forecasts (less responsive forecast to actual demand)
o Higher values of alpha, the more closely the forecast follows the actual demand = small values of n in the
weighted average method  it means also that you are always late
 low value of alpha = high value of n = smoothing forecast
 high value of alpha = low value of n = reactive forecast
 Disadvantage: lagging of the model in changes in demand (increase and decrease)
 When the casual factor/component in the demand is very high (high level of standard deviation due to uncertainty), in
that case you want a smooth forecast. A reacting forecast would be always delayed.
5
  If the variability in the demand is due only to casual variations = low value of alphas or high values of n (smooth
forecast). But when you know that the variability of the demand is mainly due to structure/systematic variability
(there is a reason behind the variability = trends, business cycles etc) is better to use reactive forecast (lower values of
n or higher values of alpha)
 Higher n = less responsive will be the forecast in respect to the actual demand = forecast smoother than actual

Mean Absolute Deviation (MAD) = is the easiest way to calculate the level of goodness
of a prediction (aka the alpha that you chose to do the forecasts).
o The lower the MAD, the better (small absolute values)
o You cannot predict always lower demand values than actual values
o 1 MAD = about 0,8 standard deviation
o Difference between what actually occurred and what was forecast

4) Linear Regression

X
Y
Z
 if you suspect that there is a trend in the demand, you can use a linear
regression
 linear regression --> the curve that minimize the square error

5) CPFR (Collaborative planning, forecasting and replenishment): model to share information to get more accurate, reliable
and fast final client demand forecasts and guesses, via software systems (web-based forecasts). It is a supply chain integration
tool that requires a high level of trust among partners in the supply chain (final forecasts are based in decision making
consensus).
Data + Trust + Organisational Structure = Constraints (CPFR)

 retailers don't want to give information about the market to the suppliers because suppliers in this was can have
bargain power

 Excel - Data Analysis --> Moving Average

o Input: actual demand (always the actual demand)
o Interval: 2 (n=2), then n=3 etc
o Check "chart output"
o When t=6 and n=2, THAT is the forecast for week 7!
o Calculate Difference between Actual demand and Forecast (A-F)
o Calculate average of A-F. This is MAD.

 Moving average: forecast is always smoother than actual demand, since we are making an average. In this case (excel file)
is better to use reactive forecast because the actual demand has a structural trend variation (the blue line presents an
increasing trend). The smoother forecast cannot react to trends!
If the demand did not have a trend (=only casual variation), the smoother forecast would be better.
 If there is a trend = reactive forecast
 If there is not a trend (casual variations) = smooth forecast
n=2 is the reactive forecast (low n)
n=4 is the smooth forecast (high n)

 Data Analysis - Exponential Smoothing

o Input range: actual demand
o Damping factor: 1 minus alpha

6
7&8: Inventory Management (Appendix G book)

 Demand mng is at the base of mng decisions, inventory decisions are at the core of it
 Decoupling points (ex. CODP): points within the supply chain where inventory is positioned to allow processes or entities
in the supply chain to operate independently. Inventory acts as a buffer to separate the customer from the manufacturing
process. Selection of decoupling points is a strategic decision that determines customer lead times and can greatly impact
inventory investment. The closer this point is to the customer, the quicker the customer can be served. Finished goods
inventory is more expensive than raw material inventory.
 Inventory = Decoupling processes  The larger the inventory, the more independent are the processes. The more
dependent, the lower the inventory. Synchronisation between processes is key.
 Keeping inventory is really expensive (30/35% of the value stocked, due to storage, insurance, obsolescence).
 How to manage inventory in such a way to reduce to a minimum the inventory while keeping responsiveness high?
 Customer waiting time = transportation time
 Inventories may include raw materials, finished products, components, parts and WIP
 Inventory management system: set of procedures and controls that monitor the stock to a minimum and establish what
level to keep, when reintegrate them, and what size orders should have.
 Purpose of Inventory:
o Maintain independence of operations (flexibility in operations)
o Meet variation in product demand (safety or buffer stock to absorb demand variation)
o Provide safeguard for variation in raw material delivery time
o Take advantage of economic purchase order size
 Inventory objectives (key decisions: when to and how many to order/deliver) and goals:
o Effectiveness: coping with changes in product demand + availability of different materials at different levels of the
supply chain, independence between phases, provide flexibility to the production plan
o Efficiency: minimum investment in capital and minimum cost managing the inventory itself
 Inventory Costs:
o Holding/carrying costs (most important!) Ex: renting the warehouse, utilities, security, insurance. Increase linearly with
the number of products you are holding. Light, warranty, internal transportation, warehouse, capital invested
o Setup costs
o Order costs (purchasing office, transportation from the supplier)
o Stockout/shortage costs (order cancellation, late delivery penalties)
o Consider costs of underestimation and overestimation

Models

1) Single-Period Inventory Model: how much to order when an item is

purchased only one time and it is expected that it will be used and then not
reordered
 News Vendor problem = z-scores
o P = 0,667  Z0,667 = 0,434
o 0,434x standard deviation of papers = number of extra papers to stock.
 Problem: if your future order from the customer will be lower of the average,
you can sell your stock the next time

2) Multi-period with Fixed Order Quantity (FOQ) + Economic Order Quantity (EOQ) = how much to order
 The ordering is triggered by inventory dropping to a specified level of inventory (ROP)
 Need to Continuously monitor and control the inventory level
 Assumptions
o Demand is stable: constant, no trend, no seasonality, only casual
o Lead time is fixed: it does not matter how much you order, but the lead time stays the same
o Price is quantity-independent
o Ordering cost is quantity independent: transportation costs
o Inventory holding cost is based on average inventory
o No backorders are allowed (all demands for products are satisfied)
 When I reach the level of R, I'll put another order to my supplier. This is a
safety stock not to run out goods
 Sometimes happen that we have to also pay shortage costs

7
 Cost Model:
o Purchasing cost: cost do not change with order size. Flat. No quantity discounts
o Ordering cost: decreases, on an annual base you make smaller number of orders.
It decreases because the process complexity decreases (also transportation costs
go down).
o Holding cost: increases, when the order size increases  how much inventory
you have on average. The more inventory you have (level), the more expensive it
is to keep it (it is less expensive to keep 500 inventories on average than 1.000)
 usually, high holding costs items will have small batches and several orders --> food, Rolex, items that request high volume of energy

 U-shaped curve Total Cost (TC)  EOQ = optimum value of order size that minimise the total cost of inventory
management (Total cost = Annual Holding + Annual Ordering + Annual Purchasing).
 FOQ  when to order?  ROP (Reorder Order Point)
 ROP: reorder point  cover the demand during lead time (you place your order in advance)  in this way we are just in
time to cover the future demand  but this is risky, I need a safety stock

𝐷 (𝑄𝑂𝑃𝑇 + 𝑆𝑎𝑓𝑒𝑡𝑦 𝑠𝑡𝑜𝑐𝑘)

𝑇𝐶 = 𝐷𝐶 + 𝑆+ 𝐻
𝑄𝑂𝑃𝑇 2

o TC: annual total cost

o DC: annual purchasing cost
o D: annual total demand
o C: purchasing cost per unit
o DS/Q: annual ordering cost
o Q: ordering size  QOPT = optimum quantity to be ordered (units) the minimize cost

2𝐷𝑆 2 (𝑎𝑛𝑛𝑢𝑎𝑙 𝑑𝑒𝑚𝑎𝑛𝑑)(𝑜𝑟𝑑𝑒𝑟𝑖𝑛𝑔 𝑐𝑜𝑠𝑡)

𝑄𝑂𝑃𝑇 = √ = √
𝐻 𝑎𝑛𝑛𝑢𝑎𝑙 𝑢𝑛𝑖𝑡 ℎ𝑜𝑙𝑑𝑖𝑛𝑔 𝑐𝑜𝑠𝑡

o S = setup cost for ordering (per order)

o QH/2: annual holding cost
o H: annual unit cost  H = iC  annual holding/storage cost per unit of average inventory. Often is a percentage of C

𝑅𝑂𝑃 = 𝑑̅ 𝐿
o R = reorder point
o L = lead time (days/weeks)
o 𝑑̅ = daily demand (D/365)

 S and H  do not depend on the inventory mng policy, but on other operation SC mng decisions
(expensive products have a higher H, and so QOPT will be lower)
 To correct the model, since demand during lead time is uncertain (risk of stockout only during lead time, aka between the
time and order is placed and the time is received), you apply a Safety Stock
o ROP has to be higher (it is not “time” but is “quantity”!)
o Probability not to go in shortage --> I have to define this probability that I want to mantain  z value
o How much risk you have will impact the safety stock. Safety stock deals with 2 risks:
 First risk: bigger/longer supplier lead time (e.g. one day of delay from supplier?)
 Second risk: demand is not constant during the lead time.

 The probability approach to calculate safety stock considers only the probability of
running out of stock and NOT how many units are short.
o How much is the variability of the demand during the lead time = standard
deviation during the lead time (L)
𝑑𝑛

𝜎𝐿 = √∑ 𝜎𝑑2
𝑑0

o How much is the likelihood not to go into shortage (z)

o Reorder point in units (R)
 If you consume the safety stock, the next time you will order before, not more  it affects only ROP and not the quantity
 If the safety stock (zL) is positive, the effect is to place a reorder sooner --> The greater the safety stock, the sooner the
order is placed.
8
 Big disadvantage: keep controlling the stock level, especially the ROP  very expensive model
 Inventory cycle counting: physically verify the inventory and where the items are, by counting them with considerable
frequency (not just once or twice a year)

3) Multi-period with Fixed Order Period (FOP)  instead of fixing the quantity you order every time, is better to control
the when your order. Fixing the when every time (fixed time windows between
orders), you can establish how much. The "how much" or the order quantities
vary from period to period, depending on the usage rates. In this case inventory
is counted only at particular times (review period). More simple, easy and less
expensive.
o Sometimes there are problems  e.g. I can order only on Monday, or only
one week per month

𝐶𝑜𝑣𝑒𝑟𝑒𝑑 𝑝𝑒𝑟𝑖𝑜𝑑 = 𝐿𝑒𝑎𝑑 𝑇𝑖𝑚𝑒 + 𝑅𝑒𝑣𝑖𝑒𝑤 𝑃𝑒𝑟𝑖𝑜𝑑

 Review period: you do not control it every time (if I order every week, the review period is one week)
 Cover period: period in which I want to cover the demand with one single order (if lead time is 3 day and my review period is
one week, I have to order items to cover one week + three days, from Monday to Thursday)
 Total covered period: have to subtract the inventory we already have

𝑞 = 𝑑(𝑇 + 𝐿) + 𝑧𝜎𝑇+𝐿 − 𝐼

o q: order quantity
o d: forecast average demand (daily/annual)
o L: lead time — from one order to another order
o T: review period
o I: inventory that we currently have
o L+T: standard deviation of the covered period = variability of
demand during this period
o z: standard deviation considered

 Differences between the 2 managing inventory models (other than what triggers the timing of the order placement):
1. FOQ size of inventory is less than FOP
o Safety stock for model FOP and FOQ: in FOP needs to be higher because the risk period is longer
o In FOP the order considers the safety stock (protect against stockout during review period itself and as well as
during the lead time from order placement and order receipt, zL+T)  FOP safety stock is higher
o In FOQ: does not consider any safety stock since it assumes continual tracking of inventory on hand
2. FOQ favours more expensive and critical items because average inventory is low. But requires more time to maintain
because every addition or withdrawal is logged. FOP is more efficient because multiple items can be ordered at the
same time and is usually used for lower-cost items.

F.O.Q. (fixed order quantity) F.O.P. (fixed order period)

How Much should we order / Order 2𝐷𝑆

𝑄𝑂𝑃𝑇 = √ 𝑞 = 𝑑(𝑇 + 𝐿) + 𝑧𝜎𝑇+𝐿 − 𝐼
Size 𝐻

When to place a new order 𝑅𝑂𝑃 = 𝑑̅ 𝐿 + 𝑧𝐿 T = review period

9
 Price Break Model = Relaxing one of the assumptions prices  quantity discount: if you order more, you will get a discount
 discount is not computed in the FOQ
 when the purchasing costs curve is not constant
 discount on quantity ordered depending on the batch
 Steps:
o Step 1: Calculate EOQ (QOPT) for every price break, from
lowest to highest (Ex: 3 prices = 3 EOQ  3 total costs you
will get with the minimum of the 3-price break)
o Step 2: verify if EOQ are feasible or not (they are within the
price break intervals)
o Step 3: calculate total costs  Where is the optimal Q? The
one that gives you the minimum cost

 Mixed models: Optional Replenishment

 Inventory Managements Indicators

• how can I measure my performance in inventory mng?
• every company can design its personal indicator!

 ABC Inventory classification (Pareto principal): divides inventory into dollar volume categories that map into strategies
appropriate for the category. Dollar volume is a measure of importance: an item low in cost but high in volume can be more
important than a high-cost item with low volume.

NOTE:

1) how can I reduce H and S (ordering costs) per unit? S and H depend on operational decisions
 S
 CORRECT: select a supplier that is next door --> less transportation
 WRONG: increase QOPT --> I'll have more holding costs and I'll reduce only total S
 H
 WRONG: reduce inventory
 CORRECT: what is based on my holding cost? On the warehouse? On the price of the good? After
understanding the single components of H, I can make efficiency on each single component

2) Total costs depend on the standard deviation on the demand --> the more the demand is fluctuating, the more you
need safety stock, that increase the costs

3) Whit a better forecast of the future, can I reduce the total costs? Not directly...  WHY NOT?

10
9: Sales and Operations Planning (SOP)

 Planning is important, but what is the horizon?

o Long term (> 1 year)
o Medium term (6-18 months)  output: Master Production
Schedule (MPS)
o Short term (1 day – 3 months)
 You always need demand forecast (without it you cannot plan)
 Capacity planning is a long-term decision: in an airport you decide
how many people to manage, in a production plant you cannot
change the number of machine station after the design
 From long-term planning (year): capacity planning, deciding
where to locate the inventory, suppliers’ agreement and
distribution/transportation systems
 Medium-term planning (monthly)  demand management and
forecasts + sales and operations planning

 the sequence is important!

 Aggregate planning (output of SOP)

 Definition: given the projected demand Ft for each period t (t=1, 2, ... T), determine the production rate Pt, the level of
inventory It and the workforce level Wt for each period which minimize the cost associated during the planning horizon
 Aim: identify the best combination of: production rate, inventory, workforce  to satisfy the demand minimizing costs
for each period (month)
 "aggregate" because is a family/group of products and customers
 It precedes the MPS
 The aggregate sales planning and operations converts the annual and quarterly business plan into the medium-term
aggregate plans on labour and other production units

 When I cannot meet the demand, I can:

 Hire / lay off (Wt)
 use inventory (It)
 out produce // subcontracting (Pt)
 use 2 turns per day (overtime is expensive) // part time (Pt)
 Mixed strategy
What is better? It depends on costs.

 Costs involved:
 Production costs (important)
 Costs associated to workforce flexibility and balancing
 Inventory holding costs (storing, insurance, taxes, spoilage, obsolescence), plus
 Backorder costs (penalties, loss of sales revenues)

 Aggregate Operations Plan (SOP)

 Objective: minimise cost of resources (labour and production) required to meet demand over a period
 Sales and Operations Planning is the process that keeps demand and supply in balance by coordinating manufacturing,
distribution, marketing, sales, and financial plans.
 It needs cooperation and team work from cross-functional departments to reach an agreement on the best course of
action to achieve the optimal balance between supply and demand.
 Help the firm provide better customer service, lower inventory, shorten customer lead times, stabilise production
rates = help the company get demand and supply in balance and keep them in balance over time (planning is the only
way to deal with dynamic and fluctuant demand).
By focusing on aggregate product and sales volumes, marketing and operations functions are able to develop plans for the way
demand will be met (fluctuation of demand due to market trends and other factors).

 Cut-and-try: to optimise the cost of a plan. Method for analysing aggregate planning problems
 example: hire/lay-off strategy  increase/decrease production capacity
 example: inventory strategy

11
 Production planning strategies
 plans for meeting demand that involve trade-offs in the number of workers employed, work hours, inventory and
shortages (Cut-and-dry)

 Output of the plan

 planned production rates, aggregate labour requirements, and expected finished good levels (Master Production
Schedule).

 Revenue Management (Yield management)

 it is the process of understanding, anticipating and influencing consumer
 the goal of yield management is to produce the best possible financial returns from a limited available capacity. It
attempts to allocate the fixed capacity of a service provider, such as airline seats or hotel room reservations

12
10: Material Requirement Planning (ERP + MRP)  subcomponent // dependent demand
 ERP (Enterprise Resources Planning) = SOP + MRP  meet demand over time
 Computer system that integrates application programs in accounting, sales, manufacturing and other functions in the
company. This integration is accomplished through a database shared by all the application programs
 Benefits: better processes, information accuracy, responsiveness through real-time information provided by the
system. Specifically, in supply chain: managing materials, scheduling machines and people, coordinating suppliers,
processing customers’ orders are all included in the ERP
 ERP supports: design of a supply chain system, future planning to meet supply and demand objectives, collaboration
between suppliers and customers

 MRP (Material Requirement Planning): used for “dependent demand” items

 technique to determine the number of parts, components and materials needed to respect a Master Production
Schedule (MRP input is MPS)
 provides a timed-schedule that specifies how much and when to order to the suppliers (or to the shop floor) the
materials, parts and components
 we need to calculate month by month the number of raw materials, sub components and parts in order to make (a
product) the Master production schedule feasible
 each component has a lead-time
 we have to decide make or buy? Usually components at the lower lever in BOM are BUY, components in the higher
part are MAKE
 You need the materials by order release and NOT by requirement

 the requirement of each components depends on the order release of its parent components

 MRP input
 MPS: the plan states how many final/finished products the company has to produce and when (gross requirements)
 BOM: identifies the specific raw materials, parts/components, lead times, production sequence of a product
 IT: inventory transactions contains the data of scheduled receipts (on order) and inventory on hand
 MRP output
 Planned Order Releases: How much and exactly when to order  it is a PROPOSAL

 MRP analysis
From top to bottom components (from independent to dependent,
planning "backwards" of lead time)
 The sub-tables have all the same layout
 Output of MRP = "Planned order release" in the sub table
 In this example, planned order receipt is equal to net
requirements = "We order just what we need"
 If safety stock is needed, the on-hand balance needs to be
reduced by the safety stock
 Net-requirements = projected available balance + scheduled
receipts are not sufficient to cover gross requirements
 Planned order receipts = amount required to meet net
requirements.
 Planned order release = planned order receipt offset by the
lead time.
 What if I have inventory? I have to transform the requirement
in NET requirement, using the inventory
 inventory transaction = scheduled receipts + proj. available
balance

 Lot sizes are the production (or purchasing) quantities used by the MRP system
 Lot-for-lot (L4L) = the system schedules exactly what is needed in each period = planned order receipt is equal to net
requirements = "We order just what we need"
o Plan orders to exactly match net requirements
o Produce exactly what I needed each week with no inventory
o Minimises carrying/holding cost
13
  Economic Order Quantity (EOQ)
 Fixed Order Quantity (FOQ)
 Fixed Order Period (FOP)
 Lot Sizing policies/problems: make planned order receipt and planned order release (and in our case net requirements
too) different. "You may order more or less than what you need due to a pre-existing inventory".
 L4L is only used to balance the fixed and variable costs that vary according to the production lot size but does NOT
consider the capacity of machines, setup costs, people and the supply chain as a whole. This model causes high setup
costs. That's why you require Capacity Requirement Planning (CRP) and MRP.

 CRP or MRP II
 MRP uses 3 main sources of information: BOM, MPS and IT
 CRP uses process plans: you find them on the shop-floor
 CRP used to check if the proposal of order release is feasible (manufacturing
process plan)
 CRP considers the REAL production capacity:
o Lead-time of production: if you need to ship 20 you can produce
them 1 day, but what if you need 2000?
 Document showing the manufacturing steps for a product or component
 Calculate the workload of each work centre, week per week. When the
workload exceeds available capacity:
o Overtime
o Add machine to perform task
o Outsource the operation
o Anticipate or postpone part of the operations
o Redefine a new execution date and re-program.

NOTE:
 One more final good has a big impact on all the chain because is linked to multiple subcomponents with the BOM.
 MRP: simulation back and forward, what if analysis --> planning the production based on the maximum capacity that is
achievable.
 Lead-time in MRP are considered fix  but it is not true
o You can calculate the average time in system

1
𝑇𝑠 =
𝜇 − 𝜆(𝑘)
o  = service rate
o  = arrival rate (load)

Finish?

14
12: Just in Time (JIT), Lean Production, Theory of Constrains
 JIT (OM philosophy)
 It is the basis for Lean Production: set of activities designed to achieve high-volume and high-quality production using
minimal inventories (raw materials, WIP and finished goods) by eliminating waste in every activity/area of production
(people management & organization, suppliers’ relationships, stocks, overproduction, process, design, manufacturing)
o e.g. excess inventory  inventory is one of them and hides other problems (machines breakdown, queues,
resource redundancy, order changes, un-reliable suppliers)
 also involves timing of production: resources arrive "just in time" to the next work station
 it focuses on having stocks of goods, providing services, and using resources only where and when needed
 The goal is to drive all inventory queues to zero, thus minimising inventory investment and shortening lead times
 When inventory levels are low, quality problems become very visible
(exposed problem hidden by excess inventories and staff!)
 JIT and Lean Production work best in repeatable and standardised
operations. Price to pay by being lean: customer service when unlikely
events occur
 Achieve high customer service with minimum levels of inventory investment
 JIT drawback: cannot rely on extra material just in case something happens
 Conventional approaches use safety stock and early deliveries against
production problems  JIT uses excess labour (overtime and part-time)
which is much cheaper than carrying extra inventory.

 Group technology: redesign of the production plant to reduce movements of parts, redesign the layout of product and
process. Aggregate products into families of products in which they are technological similar (similar technological features)
and the processes required to make the parts are arranged in a specialised work cell. The aim is to:
o Reduce unnecessary movements relating to each product (and waiting times)
o New layout, reducing material handling and improve flow of product
o Reduce inventory and number of employees required

 Waste in operations: looking a machine run, waiting for parts, counting parts, overproduction, moving parts over a long
distance, storing inventory, looking for tools, machine breakdown, rework

 Lean Supply Chain Design Principles:

1) Lean Production Layouts 2) Lean Production Schedules 3) Lean supply Chains
a. Group Technology a. Uniform plant loading a. Specialised plants
b. Quality at the source b. Kanban production control system b. Collaboration with suppliers
c. JIT production c. Determination of number of Kanban needed c. Building a lean supply chain
d. Minimised setup times

 JIT: demand-pull logic  KANBAN:

 Not keeping inventory and not producing the EOQ
 Usually is the inventory (when low) that pulls the demand. But in this
case is the customer that pulls the demand when making an order.
Reduce in advance and then supply/satisfy the demand. The system
pulls materials based on need.
 To enable this pull process to work smoothly, lean production
demands high levels of quality at each stage of the process, strong
vendor relations, and a fairly predictable demand for the end product.
 You make stock and maintain that stock. If the customer asks for parts
you can send them instantaneously and then the customer demand
will PULL your production to cover again your inventory.
 Pull logic / Kanban: the assembly line use inventory, they will empty
the box of parts and will give the withdrawal Kanban to the production (production Kanban)

 Kanban Production Control System.

Kanban pull system is an inventory or production control system that uses a signalling device (Kanban) to regulate and control
JIT production flows. The authority to produce or supply additional parts comes from downstream operations.

15
Part A storage = container with part A
Withdrawal Kanban = it's a tag

 Designing and setting up a Kanban control system requires to

establish the number of Kanban containers/cards. And EACH
container represents the minimum production lot size to be
supplied. The number of containers, therefore, directly controls the
amount of work-in-progress inventory in the system.
 Accurately estimating the lead time needed to produce a container of parts is the key to determining the number of
containers. This lead time is a function of the processing time for the container, any waiting time during the
production process, and the time required to transport the material to the user. Enough Kanban are needed to cover
expected demand during this lead time plus some additional amount for safety stock
 A Kanban system does NOT produce zero inventory, it controls the amount of material that can be processed at a
time (= the number of containers of each item). The Kanban system can be
easily adjusted to fit the current system because card sets can be easily
added or removed from the system. (ex: add/remove additional container
of material with the accompanying Kanban cards)

1) Lean Supply Chain Principles:

 Lean Suppliers are able to respond to changes
 Lean Procurement: automation
 Lean Manufacturing + Lean Warehousing + Lean Logistics + Lean Customers
2) Respect for people (JIT is also HR management)
 Fair Salary
 Collaboration of management with company Unions
 Internal CSR
3) Total Quality Control (TQC), quality at the source, focus on quality! (six , …)
 Direct responsibility of the worker for the quality of their output
 Measurement with statistical quality control
 Strict compliance requirements
 Automatic inspection
 seeks to eliminate causes of production defects
4) Quasi-Zero Inventory:
 Production is "pulled" by the demand
 Reduce lot sizes
 Collaboration with suppliers to:
o Reduce lead time
o Frequent deliveries
o Plan resource needs
o Quality goals

 Quasi-zero inventory
 traditional: ordering cost increase with order size (optimum value), and set up cost is treated as a constant. Work on
the green line: reduce the ordering cost (ex: selecting supplier that is close, so the transportation cost is lower).
 JIT/Kanban: work on the single unit order cost. Setup cost is significantly reduced and the corresponding optimal order
quantity is reduced (EOQ is lower than Traditional one) (to speed up Set-up times go to page 363)
 Setup costs (it is the order cost but for the production)  in order to do that, they did operational improvement to
decrease the set-up cost
 JIT  don't accept high fix costs

JIT  integrated set of activities designed to achieve high-volume production using minimal inventories parts that arrive
exactly when they are needed. Goal: Reduce setup costs to permit smaller lot sizes.

16
 Theory of Constraints (TOC)  synchronous manufacturing
 MRP: uses backward scheduling approach because you start with the finishing product requirements. Is oriented
toward meeting due dates and maximising the use of capacity
 JIT: uses Kanban and work-in-progress (something to pull, ex: pulls material based on need). Does not allow much
flexibility for changes particularly when capacity is tight
 TOC: uses forward scheduling (synchronised manufacturing to achieve firm goals). More flexible and focused on
maximising flow through the system while minimising cost

 Types of Constraints:  Issue of TOC

 Physical (operations and logistics) • Statistical fluctuation: series of machines working
o Internal (machines capacities) in tandem, trying to understand which is the
o External (sales, materials). bottleneck! Deviation will kill you, so it is better to use
 Managerial (organisational) the pace of the bottlenecks.
o Rules, norms • How can I reduce fluctuation?
 Behavioural (human) 1) perfect maintenance of machines
o Beliefs, assumptions 2) material handling to fully satisfy the machine
3) more flexibility in the process workflow
 Performance Measures for the GOAL achievement:
1. (+) Throughput (rate at which money is generated by the system through sales)
2. (-) Inventory (all the money that the system has invested in purchasing things it intends to sell)
3. (-) Operating Expenses (all the money that the system spends to turn inventory into throughput)
To succeed, a company must simultaneously increase throughput, reduce inventory and reduce operating expenses.
 If I have two processes with an equal throughput and different standard deviation, which one is better to have first? It
depends: if before the "bottlenecks" I don't have any waiting time, is better to have the bottleneck before

 Why unbalanced capacity is better than balanced? (The Goal notes)

 Stochastic depended events
 Balanced capacity leads to too many problems because every
resource becomes dependent to every other. Since statistical
fluctuations are inherent in any process, perfect balance leads to
disruptions
 Focus on the type of Constraints: these are the limits to
performance achievements (ex: bottlenecks)
 Managing the flow through the bottlenecks is essential. Saving time
on a bottleneck is the only way to increase throughput
 Productivity: a measure of how well resources are used.

 To successfully apply TOC:

 Resolve conflict with conventional accounting and marketing/sales
 Traditional cost accounting is based on a goal of fully utilising
resources and minimising cost. TOC on the other hand maximises
profit by improving throughput.
 instead of moving the operations, you need to balance the variation
 Sales and operation planning: lead time and capacity balancing
 Major goals in fixed capacity = minimise delay of a set of orders to
meet the due dates (reduce lead time)
 Materials and requirements planning: the goal is to lower ordering
and holding costs

 JIT vs Theory of Constraints

 JIT = objective is to reduce/minimise waste in MANY activities and to do that, it minimises inventory. Stock is produced
only when the machine needs it
 TOC synchronisation between bottlenecks and demand + have inventory at bottlenecks (have inventory close to the
machine)
 Wrong things to say: JIT means not inventory (production pull ask to have inventory); it means minimizing inventory
by eliminating every type of waste (reduction of purchasing cost? by choosing closer supplier to have good
relationships and cheaper transportation)

17
13&14: O&SCM Processes
 Why not automate (Ford VS Mazda example)? Lesson learned:
 Don’t automate, obliterate (automate only necessary
activities). Automation is not always the solution  streamline
the process first. Eliminate inefficiencies, do not make them
bigger with automation.
 Process orientation (solution is not in the office, but in the
process)

 Process: set of activities performed by an organisation that

takes inputs and transforms them into outputs that create value
 Cycle time: average time between completions of successive
units in a process

 Process Orientation
 Process-oriented organisations continue to be structured in functions
but the focus lies in the strategic and operational streamlining, re-design
and optimisation of business processes
 It's not an organisational model but a management approach (which
leads to re-envision the organisation).
 It does not replace the functional structure, but they overlap
 The resources belonging to a given business function cooperate with
those of other functions to achieve a common goal, the goal of the process
 The resources belonging to a given business function cooperate with those of other functions to achieve a common
goal, the goal of the process
 The success of process orientation depends precisely on the ability of coordination between the functional areas of
the company in such a way to ensure that business outcomes are something more than a simple sum of parts

 Internal customer and process owner

 The purpose of each process is to create added value throughout the value chain (from design to after sales)
 This value chain is composed of linking rings client/supplier within the company for which each resource operating in a
process interfaces with an internal client to which it provides the result of its activity
 The objectives of the process (external or internal client satisfaction) are defined by the process owner (responsible
for the process) which coordinates the action of the resources, presides over the activities and identifies the
performance indicators of the process.

 O&SCM process
1) Forecasting, commercial agents’ requests, received orders
2) First attempt MPS
3) Rough cut capacity planning (RCCP)
4) Authorized MPS
5) MRP
6) MRP II / CRP
7) Shop floor control (SFC)
 communication, collaboration and coordination of
activities belonging to different functions.
 Process Flow Analysis (evaluate and improve process performance)
o Activities of the process "Production management": ensuring the availability of space-time materials, from
suppliers to the warehouse of finished products downstream of the factory along the entire chain of
production
o Supplier relationship management: aims to ensure (not perform) the business logistics from suppliers
(suppliers’ selection, contract management, evaluation etc)

AS-IS configuration VS WHAT-IF configuration

Exercise of airport

18
 Strategic alignment
 Business strategy: defines the goal of every corporate division or business unit in terms of competitive priorities that
will be offered to each market segments that will be served (strategic positioning).
 Functional strategy: defining the objectives of the three major business functions:
o Marketing strategy: identifies customers that the business unit wants to serve, the products and the 4 P that it
has to offer in response to customer needs
o Operations strategy: defines the design, planning and management processes through which the business unit
will provide services and products to customers
o Finance strategy: acquires and allocates resources to various processes of the business unit
 Performance
o Financial measures: relative to the difference between the value of the goods or services offered to the
customer and the cost of production/sale
 Absolute measurements (income, expenses, net income, profit) & measures relating to resource use
(ROI, turnover ratios)
o External performance measures, relative to customer satisfaction (process output)
 Cost of product (aka price), delivery time, product variety, quality of product
o Internal performance measures, relative to internal efficiency of the process
 Cost of process, flow-time process, process flexibility, process quality
 The operations manager has to translate the customer expected values (external measures) into the appropriate
measures of internal process
 Output of operation has a higher value of the input

 Process flow analysis (evaluate and improve process performance): transform input in output
1) Input and output
a. inputs of a business process are tangible or intangible units that flow from outside to the process (raw
materials, work in progress, energy, information)
b. output of a business process are tangible or intangible units that flow from the process towards the external
environment (finished products, materials)
2) Flow units: the entity that is processed (ex: in the airport example, is the passenger)
3) Network of activities and buffers
a. basic activities are the simplest form of transformation of inputs into outputs
b. buffer stores flow units that have already been processed by a task and are waiting to be processed by
another task (inventory and storage of the flow unit, increases and decreases)
4) Resources (tangible assets that can be capital) (buildings, land, production systems, machines, information systems,
etc..) or labour (human resources)

 Internal measures of a process performance (internal measures of every company are related to their market):
 Flow time
o total time necessary to transform input to output (processing time + waiting time in the buffers)
o very useful information for a manager who has to promise a delivery
date to the customer
 Throughput (flow rate)  it is ALWAYS less then capacity
o number of units flowing through a specific process for each unit of time
(capacity of the process)
o it can change over time
 Work in Progress (WIP)
o number of flow units (entities) present within the boundaries of the process
o it can change over time
Direct influence on the cost and response time and are directly influenced by the
flexibility of the process and its quality (- Cost of the process; - Quality; - Flexibility; -
Responsiveness).

 Little Law
 WIP work in progress, T flow time, X throughput
 A process is stationary when the steady-state average rate of input stream equals that of output. In such case, the
flow rates (in-flow, out-flow) are called average throughput of the process. In this case, by the known Little's law, it is
possible to determine the average value of the WIP.
 average value of WIP in steady state (average throughput of the process)
 Little Law limitation: process must be operating for a while, is stable and there's inventory at every step.
19
  Inventory turnover ratio & Little Law: number of times we are filling the inventory within a month

𝑆𝑎𝑙𝑒𝑠 𝑇ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑋 1
𝐼𝑛𝑣𝑒𝑛𝑡𝑜𝑟𝑦 𝑡𝑢𝑟𝑛𝑜𝑣𝑒𝑟 𝑟𝑎𝑡𝑖𝑜 = = = =
𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑖𝑛𝑣𝑒𝑛𝑡𝑜𝑟𝑦 𝑊𝐼𝑃 𝑊𝐼𝑃 𝑇

Cost Quality Responsiveness

X (throughput)

WIP (I)

 Depending on the type of the service, different considerations on how X, T, and WIP impact Cost, Quality and
Responsiveness of the processes.

 Examples:
 Example 1: high levels of WIP - The main effect of high levels of WIP is on costs (holding cost), but in turn we are able
to be more responsive (cycle times diminishes and we are able to reduce delivery time). Contrarily, in service
operations, when the flow unit is a customer (for example a passenger in the airport), the main effect of having high
WIP is on the flow time (more precisely on the waiting times). So, the quality of the service (expressed by the
satisfaction of the customer) decreases.
 Example 2: Flow Time - when the flow unit is a physical good (for example a product) waiting time means high
inventory. This means costs (holding cost), but also efficiencies (resources are not under-utilised because of missing
components). In service operations, the waiting time component of the flow time is something unwanted because it
will decrease customer satisfaction. So, the quality of the service (expressed by the satisfaction of the customer)
decreases.

NOTE:
 Inventory in services is BUFFER
 Inventory is in every office  e.g. documents to be processed
 In services it is very difficult to decide to what bucket put all costs

20
Wrap-Up

 Your system is perfectly design to give the result that you get
 Changing the design imply to have a different behavior
 Supply chain is not a chain but is network force
o Supply chain are dispersed  components from all countries
 Operations are the building block of supply chain
 Supply chain is disintegrated: numerous products with short life cycle and many players in different chains
 Pascal Law & Little Law
 Bullwhip and clock-speed effects
 JIT, TOC and ERP(MRP)  collaboration
o ERP: long term planning & information processing capabilities
o JIT: philosophy of waste & POOGI & Kanban system
o TOC: concentrate on constraints in short and long term & preserve flow through constraints & abolish
accounting based performance measurement
 There is no one size fits all
 Best Production and Inventory Control Systems (PICS) requirements:
o Information system requirements
o Long-term and medium-term planning function
o Capacities and priorities logic
o Shop floor (short-term) control and tracking function
o Work in progress (WIP) control logic
 Best PIC System (integration of all processes)
o ERP (SOP-MRP)
 Long-term planning and reporting
 Information processing capabilities
o JIT:
 Philosophy of waste elimination + POOGI
 Simple shop floor control (Kanban system)
o TOC
 Concentrate on constrain tin short as well as long-term
 Preserve flow though constraints
 Abolish accounting-based performance measurement

21