DM202 Sem2 Wk4 Prod Planning Control no code.pptx

Manufacturing Management DM202 Semester 2 – Production Planning Colin Andrews: [email protected] Example of factory operations ... https://youtu.be/yROGg3_vHBc Agenda  Time Horizons  Forecasting  Capacity Planning  Matching to real demand  ‘Economic’  Re-order  ‘Takt’ ba...

Manufacturing Management DM202 Semester 2 – Production Planning Colin Andrews: [email protected] Example of factory operations ... https://youtu.be/yROGg3_vHBc Agenda  Time Horizons  Forecasting  Capacity Planning  Matching to real demand  ‘Economic’  Re-order  ‘Takt’ batching level / point time Time Horizons  Short Term A period in which all capacity is fixed and resources cannot be flexed  What can we do with what we’ve got ‘now’  Typically  a week / month Long Term A period in which capacity can be completely changed and new resources can be added (old resources replaced)  What will we need in the ‘future’  Typically  > 1 year Medium term  In between these extremes  What  How might we need to achieve can we flex current resources to achieve this  Typically quarter years or half years Time Horizons  These are different for every business e.g.  Long term  Electric Utility → 30 years  Mobile phones → 3 years  Supermarkets → 12 months  Short term  Electric Utility → 1 hour  Mobile phones → 1 month  Supermarket →1 day Forecasting and Capacity Planning Forecasting and Capacity Planning  Aim  To  introduce the concepts of forecasting and capacity planning Reading  Slack  et al textbook, Chapters 10, 11, 12 Learning Outcomes  Ability to describe the importance of forecasting  Ability to calculate capacity related information and data Why do organisations forecast sales? ………many supply decisions must be made ahead of demand, e.g. production planning, resource planning, sourcing, distribution planning, etc Despite the advances of supply chain management and the emergence of concepts such as agile, lean, just-in-time, etc……. It takes time to: • Source raw materials, produce and distribute products; • Recruit new staff, find new sources of supply, launch an advertising campaign; • Build new warehouses in new locations, develop new products or technologies, etc. We need to forecast future sales so that we can carry out: • Demand and Inventory planning • Capacity, production and distribution planning • Financial planning • Business planning Planning Horizons: How far into the future do I need to forecast? Level Reviewed Horizon Business Unit /Categories annually 2 – 5 yrs Marketing Product Family/Brand Quarterly / monthly 12 – 18 months Production Planning Individual SKU Monthly / weekly LT - 9 months Stock Replenishment Individual SKU Weekly/daily Lead-time Purpose Business Planning Strategic Tactical Operational Forecasting is not easy - the only certainty about a forecast is that it will be wrong ! “Two-thirds of all sales forecasts have a margin of error that exceeds 25 %” “More than 10% of sales forecasts have a margin error greater than 70%” Source: Sales Benchmark Index Demand planners, like weather forecasters rarely get credit for doing a good job, they’re only noticed when they get it wrong (which is most of the time!) Poor forecasts create imbalances in supply and demand resulting in…. Either ? Excess Inventory: increased capital employed, obsolescence, relocation costs, clogged up supply chain, discounting, etc Or ? Poor Availability: late /emergency deliveries, lost sales, lost profits & lost customers ! Why is forecasting difficult? Unexpected ‘events’ New product introductions Promotions Competition Product proliferation Customers !! Short product life cycles Advertising campaigns “…the volatility of markets ensures that the forecast will be wrong. Whilst many forecasting errors are the result of inappropriate forecasting methodology the root cause of these problems is that forecast error lead-time increases as ..............increases” A ‘Good’ forecast should form the basis of a more consistent match between supply and demand Cost of Lost Sales Bad Poor Availability Increase supply Increasing cost Better Best Source: Colin Sheppard ? Decrease supply Cost of Excess Inventory High Waste Bad How can we improve the match between supply and demand? Communicate & Collaborate: Cross-functional (S&OP) Cross-organisational (VMI, CPFR) Build a Robust Baseline Forecast: use best-fit algorithms and causal models Manage the Lead-time Gap: Reduce Lead-times and become more Agile The need for Integrated cross-functional planning SUPPLY CHAIN OPERATIONS MARKETING & SALES Demand Creation Demand fulfilment Demand fulfilment Demand Creation Integrated Planning Alignment of demand & supply planning processes across functional boundaries The need for integrated cross-functional planning Mid 1990s – Volvo held an excessive inventory of green cars. • What did they do? • What would be the implication of this action? Source: Lee (2001) What is needed? Synchronised Planning cycles across functions Demand Planning Sales Sales Plan Plan Supply Plan S & OP* meeting 4th week S&OP meeting *Sales & Operations Planning Supply Planning Principles of Forecasting • Common Features of Forecasting Models: – Forecasts are rarely perfect – Forecasts are more accurate for groups or families of items rather than for individual items e.g. Luxury SUVs rather than Audi Q8s – Forecasts are more accurate for shorter rather than longer time horizons (for same level of item) Steps in the Forecasting Process 1. Decide what to forecast 2. Evaluate and analyse appropriate data 3. Select and test the forecasting model 4. Generate the forecast 5. Monitor forecast accuracy Forecasting Methods Qualitative Methods Quantitative Methods Characteristics Based on human judgement, opinions; subjective and nonmathematical Based on mathematics; numerical in nature Strengths Can incorporate latest changes in the environment and ‘inside information’ Consistent and objective; able to consider much information and data at one time Weaknesses Can bias the forecast and reduce the forecast accuracy Often quantifiable data are not available. Only as good as the data on which they are based Forecasting Methods  Qualitative Methods  Executive opinion  Market research  Delphi method  Quantitative Methods  Simple mean  Simple moving average  Weighted moving average  Exponential smoothing  Seasonal indexes  Linear regression  Multiple regression Needs relevant history / data Some common Statistical (Projective) Techniques Increase value of n to smooth out Moving Average lumpy or volatile demand All values given equal weight Simple Exponential Smoothing (SES) Smoothes level only More weight given to most recent data Double Exponential Smoothing (DES) Smoothes level and trend Holt’s Exponential Smoothing Smoothes level and trend Croston’s Exponential Smoothing Smoothes level and intervals between +ve demand Holt-Winter’s Exponential Smoothing Smoothes level, trend and seasonal factors Time Series Decomposition Multiplicative or Additive Decomposes underlying patterns and projects into the future. NO smoothing Forecasting Methods  Factors Influencing the Selection of Forecasting Models:  Amount and type of data available  Degree of accuracy required  Length of forecast horizon  Data patterns present Links Between Plans Forecasting Marketing Plan Aggregate or Production Plan Aggregate or Capacity Planning • Details the aggregate production rate and size of the workforce, which enables planners to: – Determine the amount of inventory to be held – Amount of overtime or ‘undertime’ allowed – Authorised subcontracting – Hiring and firing of employees – Back ordering of customer orders Capacity Planning Definitions:  Capacity  the  maximum output rate that can be achieved by a facility Capacity Planning  the process of establishing the output rate that can be achieved by a facility Capacity Planning Demand Machines Product People Customer Material Money How to reconcile customer demand with the supply of input resources Determining Capacity  Start with forecast demand  Number  e.g.    of top level items per period (independent demand) bicycles Use the BoM to work out the demand for lower level items  Sub assemblies and components (dependant demand)  e.g. wheels (2 per bike demand) … and the route cards to identify how much time is required on each operation  Frame welding  Wheel assembly etc. Either  Calculate the number of machining / assemble / work centres required to meet demand  Compare the time required of each available work centre against demand to see if it is possible Capacity planning Issues:  Too much capacity  Excess costs  Too little capacity  Order Backlog  Idle facilities  Lost  Idle workers  Cannot  Idle equipment sales satisfy customer demands Links Between Plans & Capacity Planning 1 2 3 4 5 6 7 8 9 10 Total Forecast (Pieces/ Week) 35 25 10 80 0 85 75 150 45 5 510 Production Plan (Pieces/ Week) 50 50 50 65 65 80 80 80 55 50 625 Capacity Requirement Plan 50 50 50 65 65 80 80 80 55 50 625 Capacity Available 70 70 70 70 70 70 70 70 70 70 700 +20 +20 +20 +5 +5 -10 -10 -10 +15 +15 n/a Week Capacity Gaps Available Capacity • Design Capacity – Maximum output rate achieved by a facility under ideal conditions – Can only be sustained for short periods of time – Achieved through overtime, subcontracting, etc. • Effective Capacity – Maximum output rate sustained under normal conditions – Include work schedules and breaks, machine maintenance, etc. NOTE: effective capacity is usually lower than design capacity Capacity Utilisation Definition: Percentage measure of how well available capacity is being used Utilisation= actual output rate / design capacity (100%) Making Capacity Planning Decisions Forecasting Capacity Capacity Cushions 1. Identify Capacity Requirements Strategic Implications Do nothing 2. Develop Capacity Alternatives 3. Evaluate Capacity Alternatives Expand large now Expand small now, with option to add later Decision Trees Rough Cut Capacity Plan   Simplify the demand forecast into a single (rough cut) measure  Could be machine hours  Could be ‘standard units’  Could be bottle neck demand Vary the plan to ensure Rough Cut Capacity is not exceeded  Pull activity forward into gaps  Make for inventory  Organise  LAST overtime RESORT – delay order Links Between Plans & Capacity Planning Week Forecast Plan (Pieces/ Week) Production Plan (Pieces/ Week) Capacity Req.ment Plan Capacity Available Capacity Gaps 1 2 3 4 5 6 7 8 9 10 35 25 10 80 0 85 75 150 45 5 Total = 510 50 50 50 50 Pulling orders 50 forward 65 65 80 Rough Cut 70 70 Measure 70 70 70 70 +20 +20 +5 +5 -10 50 +20 65 65 80 80 80 55 80 80 Total = 625 55 50 70 70 70 Plan doesn’t -10 -10 work 50 70 Total = 700+15 +15 Capacity Plan & Utilisation  Is it good to plan for 100% utilisation?  NO!  Leaves no flexibility to respond to ‘within plan’ situations  Capacity is an assumed number – depends on many factors  ‘Right’ plan level should be based on experience (and challenged)  Return to this later in the course! ‘Economic’ Operations  Cannot always afford to make to stock  Inventory is costly to hold  Changing production hurts utilisation  Businesses driven towards operating in batches  What is the ‘right’ batch size  To Order: Economic Order Quantity  To make: Economic Batch Quantity  When to order: Re-order point, Re-order level Economic Order Quantity (EOQ)  Costs associated with holding stock (Ch)  Costs associated with ordering stock (Co)  Working capital cost  Administration  Storage cost  Transport  Obsolescence risk cost costs  Opportunity discounts costs cost of missed EOQ  Q = Maximum Inventory (Q/2 = average inventory for smooth demand)  D = Number of deliveries per period  Total Cost of Inventory (Ct) = Ch + Co  Holding costs can be estimated (holding cost per item x average number of items) = Ch x Q/2  Delivery costs per unit can be estimated (cost per delivery x deliveries / inventory quantity) = Co x D/Q  So Ct = ChQ/2 + CoD/Q EOQ  Can an ‘optimum’ be found for Ct = ChQ/2 + CoD/Q?  As inventory quantity is lowered, inventory ordering costs rise.  The minimum is when the rate of  change of Ct with Q is zero  i.e. dCt/dQ = 0 EOQ  So ... dCt/dQ = Ch/2 – CoD/Q2 = 0  Solve for Q ... Q = (2CoD/Ch)½ EBQ  A similar approach can be used to set an Economic Batch Quantity: D = (Market) Demand rate R = Replenishment (production) rate Re-order Level  Assume inventory is depleted at a steady rate  Further, there is a dependable ‘lead time’ for deliveries  When do I place my EOQ order?  When: stock level = use rate*lead time  Known as Re-order Level  (or Re-order Quantity) Re-order Point  Issues with Re-order Level:  Order period is unknown if the use rate varies  Stock outs may occur if the lead time is longer than expected or the use rate is greater than expected  Requires   continuous checking of stock levels Or highly accurate stock management Re-order Point inventory management is an alternative Re-Order Period  Stock ordering reviewed at a set period  Quantity ordered can be:  EOQ  Sufficient  Some  to reach target stock level alternative quantity Benefits  Ordering happens to a set schedule ‘this months order is for ..’  Ordering is flexible and can respond to changes Real World ‘Adaptations’  Include Safety Stock  Use forecasting ranges  Focus on uncertain items  It’s a full-time job! Criticism of ‘Optimum’ approaches  There are a number of criticisms of the EOQ / EBQ optima:  The assumptions in the analysis are too simplistic  The real costs of stock in operations are not fully visible  The models themselves are descriptive and should inform decisions rather than prescriptive (give ‘the’ answer)  Cost minimisation is not necessarily the best approach for inventory management  Alternative ‘optimum’ approaches are available ... Takt time  From the German word for a Conductor’s baton  Definition:  Drumbeat cycle of the rate of flow of products  Regular, uniform rate of progression of products through all stages from raw material to customer  Objective  Synchronise  Set all work flow with customer demand a regular pattern of activity Takt Time Takt Time = Net available time to work / customer demand Takt Time Example • 5 working days/week • 2 shifts • 8 hrs per shift • 1 hr allowance per shift • 120 parts/week demand Takt Time = Net available time to work / customer demand Takt time = (2x (8-1)x5)x60/120 mins = 4200/120 = 35mins All customer focussed operations should be carried out in 35 minute cycles Work Content Smoothed  Before Smoothing After Smoothing Work Content Smoothed Before Smoothing to Takt Time Takt Time Takt Time After Smoothing to Takt Time Operator Flexibility Finished Goods  4  1 I   Raw Materials Finished Goods   6 1 Raw Materials   H 3 2   G 4  1  A B C I H G 5 2 A   F 4 3 B   5 1 C 3 3  2  2 E  3 E 3 Operators 1  D F 4   2 D 2 Operators