ISy E 6203 Inventory vs Transport John H


























































- Slides: 58
ISy. E 6203 Inventory vs Transport John H. Vande Vate Spring 2012 1 1
Agenda • Team Presentations – Financial Statements • Transport & Inventory – How’d you calculate? – What would you do next as VP of SC? • Improving the balance between transport & inventory 2 2
Our Company • Transportation: – $435 K • Inventory: – Green Bay: $900 K – Indianapolis: $600 K – Denver: $50 K – Stores: $155 Million – Pipeline: $1. 654 million • How to reduce inventory at stores? 3 3
First Step • Use full-truck load rates, but don’t fill the trucks • Treat all stores the same • What happens if we ship half a truck load instead of a full truck load? • More generally? If we ship a quantity Q? 4 4
Total Cost Formula • Transportation Costs – D = Annual Demand at the Store – A = Transportation cost per shipment – Annual Transport cost: AD/Q Note that’s because we pay for the full truck regardless of what’s on it. • Inventory Costs at the Store – h = Inventory holding cost % – C = Value at COGs – h. CQ/2 5 5
Total Cost rises quickly as shipments get very small – move to a different mode or… Total cost typically flat so anything close to minimum is good 6 6
Total Cost Formula • How does changing the shipment quantity Q affect pipeline inventory? 7 7
Minimize Total Cost • • Total Cost (Q) = AD/Q + h. CQ/2 Simple incremental analysis: – – Look at increasing Q by small amount ε Incremental change in Inventory • – • • +h. Cε/2 inventory impact to order ε more Incremental change in Transportation The increase in inventory is independent of Q The decrease in transportation is declining in Q Increase Q until the incremental transportation savings = incremental inventory increase, beyond that Total Cost increases again 8 Recognize this? 8
The EOQ • Increase Q until incremental savings in transportation = incremental increase in inventory • AD/Q 2 = h. C/2 • The Economic Order Quantity • Calculus: 9 9
The EOQ • • Achieving Balance AD/Q 2 = h. C/2 => AD/Q = h. CQ/2 Annual Transport Cost = Annual Inventory holding cost If Transport Cost > Inventory holding cost, shipments are too large? Too small? • Our Company: – Transport Costs – Inventory Holding Costs at Stores • Inventory Holding Costs at Stores • Conclusion? 10 10
Inventory at Source? • What about inventory at the Source? Green Bay, Indianapolis or Denver • Our Total Cost formula ignored inventory at the source • If we ship Q to every store, what will the Cycle Stock at Green Bay be? Q/2 • How does this affect our analysis? It changes the EOQ ever so slightly as we will see. 11 11
Inventory at Source? At the stores How does this affect our analysis? • Total Cost (Q) = 100 AD/Q + 100 h. CQ/2 At the + h. CQ/2 plant = 100 AD/Q + 101 h. CQ/2 • So we want 100 AD/Q 2 = 101 h. C/2 So the adjustment to the EOQ really is slight 12 12
Our Company: Challenge 2 • Determine the appropriate shipping quantities for Our Company • Adjust the financial statements to reflect changes in transportation and inventory • Discuss the impacts • What would you, as VP of SC, do next? • Two Parts 2 A and 2 B • That’s due Tuesday, Jan 24 th 13 13
Challenge 2 A • Do the simplest thing first, if it works and elaboration is worth the effort and added complexity, then elaborate • If we ship a different quantity Qi to each store, what is the inventory impact at the source? • Let’s first calculate the appropriate shipping quantities (and the financial impacts) if we treat all stores the same. Think more practically about what to do next. 14 14
Challenge 2 B • Elaborate on the simple model • Calculate appropriate shipping quantities (and the financial impacts) • Balance between impact and how practical it is to implement • Treating each store separately is not practical 15 15
Treating different stores differently • Two Steps: – What impact will this have on inventory at the source? – Practicalities • Does it make sense to have a separate quantity for each store? How would you implement that? What’s it worth? • Is there a practical way to achieve most of the savings while avoiding most of the complexity? 16 16
Impact at the Source • Green Bay ships quantity Qi to store i, – What’s the inventory impact at Green Bay – What’s the appropriate quantity to ship • Think about just one store. • All we have to do is figure out the Cycle Stock contribution at Green Bay 17 17
Cycle Stock at Green Bay • Worst Situation: Synchronized Shipments – Ship to all 100 stores simultaneously – Cycle stock at Green Bay: Qi/2 • Assumption: Green Bay will allocate production to minimize Cycle Stock • Rotate full attention to one (a small number of) store at a time, staggering shipments as much as possible. • We will simply assume we can stagger shipments perfectly • What’s the Cycle Stock? 18 18
Cycle Stock • • Look at shipments to 1 store They are of size Q How long to build up that shipment? P = Rate of production Area = Item-days of inventory Q Q/P 19 19
Green Bay Cycle Stock • item-days of inventory at the Green Bay accumulated for each shipment to Store #1 if the shipment size is Q Q Q/P Area = item-days of inventory 20 20
Green Bay Cycle Stock • How many shipments to Store 1? D/Q • Total Item-days of inventory at Green Bay for Store 1: Q Q/P Area = item-days of inventory 21 21
Green Bay Cycle Stock • Total Cost between Green Bay & Store 1 • Recognize this? 22 22
Economic Production Quantity • EPQ 23 23
Green Bay Cycle Stock • If we ship a different quantity for each store, Qi for store i, what’s the cycle stock at Green Bay D is the annual demand at a • Store i contributes store. • So the total is • More generally • Or half the demand-weighted average shipment quantity 24 24
Discussion • So the basic EOQ is good enough here – Many, similar customers • EPQ more relevant if there are few customers or large customers so that D and P are comparable • We will return to this 25 25
Frequency vs Quantity • Often more practical to set a regular schedule for replenishment, e. g. , weekly or semi-weekly… • n = Shipments per year The cycle inventory at the plant • Q = D/n (careful with units) • Total Cost (n) = AD/Q + h. CQ/2 26 26
Frequency vs Quantity • Total Cost (n) • So that the best value for n is 27 27
Practical Approach • Simplify by grouping Stores into Frequency Bands • For example: – – Band 1: Stores served once per 2 -months Band 2: Stores served once per month Band 3: Stores served twice per month Band 4: Stores served 3 times per month • The appropriate choices will depend on the “economics” • Observations: – Where will Band 1 stores be? – Where will Band 4 stores be? – Why? 28 28
Where’s Band 1, Band 4? • Normally think of the EOQ (EPQ) as – Given a fixed transportation cost defined by distance, find the appropriate frequency • Turn that around – Given a fixed frequency, find the appropriate transportation cost and so distance The higher the frequency the closer the stores should be 29 29
Dividing line between Bands • Dividing line between Band 1 and Band 2 – Band 1: Stores Served once per 2 -months (6 times per year) – Band 2: Stores Served once per month (12 times per year) • Find the fixed cost (distance) A so that the total cost at n 1 = 6 and at n 2 = 12 shipments per year is equal That’s about 975 miles Stores closer than 975 are better served 12 times per year. Stores farther are better served 6 times per year. 30 30
Changing Modes • Not necessary to pay full truck load rates for less-thantruck load shipments • Can use LTL services – – – YRC Worldwide Fedex Freight UPS freight ABF Freight System Old Dominion Freight Lines Estes Express Lines • An excellent guide to using LTL services is available at LTL Training Manual Module 1 LTL Training Manual Module 2 LTL Training Manual Module 3 31 31
LTL • LTL – Less-Than-Truckload: Carrier organizes transport via consolidation terminals • Groupage: Similar concept, but organized by a forwarder who contracts with a carrier for transport.
LTL • LTL
LTL Rates LTL Cost Shoulder Constant rate Rate per CWT 34 34
Observations • Economies of Scale – Although the cost of larger shipments is higher, the rate is lower – This is generally true across modes: Parcel – LTL – Rail Car … • Under LTL we only pay for the portion of the vehicle we use (but at the higher rate) 35 35
Observations • LTL rates depend on – Origin – Destination – Weight (in hundred weight CWT) – Class of freight • Includes density, value, … • So there are lots of LTL rates • Rating Engines – Czar. Lite, (SMC 3 Southern Motor Carriers Rate Conference) – Radical Tools (MARS 500 Middlewest Motor Freight Bureau) 36 36
Radical Tools • Available in the labs • Excel Add-in – Includes • • Zip code look up Zip code to City, State 3 -digit zip to Zip Road distance between Zips Find the closest point in a set of Zips Find the cheapest (in terms of LTL rates) point Rate LTL shipments Rate Parcel shipments 37 37
$/CWT Specific Example TL Cost LTL Min Charge Shoulder Includes a 24. 2% fuel surcharge After the 70% discount Pipeline impact 38 Download LTL Rate Worksheet from T-Square 38
Question • How do we determine the appropriate shipment size trading off LTL and TL rates and inventory carrying costs? • A first step: – Between 2, 001 and 4, 346 lbs the rate is $98. 70 per CWT (before the 70% discount) or $29. 61 after the discount – How do we think about how much to send in each shipment under this rate? – How is this different than FTL rates? 39 39
Question • A first step: – $29. 61 per CWT vs $4851. 68 per TL • What affect does making smaller shipments have on transport cost at the per CWT rate? • What affect does making smaller shipments have on inventory cost at the per CWT rate? • What shipment size is appropriate at the $29. 61 per CWT rate? 40 40
Question • How do we determine the appropriate shipment size trading off LTL and TL rates and inventory carrying costs? • A second step: – Between 4, 347 and 5, 000 lbs the rate is $1287. 15 regardless of weight. – How do we think about how much to send in each shipment under this rate? This is probably over-kill except perhaps at the top shoulders, e. g. , 13, 820 – 30, 000 lbs 41 41
Challenge 3 • Determine the appropriate mode and shipping quantity from Green Bay to Store #2 in Los Angeles. • Briefly describe thinking behind your calculations 42 42
Challenge 4 • Roll this strategy out to all Stores. Determine the appropriate mode and shipping quantity to every store • Remember to keep it practical. • Briefly describe thinking behind your calculations • What are the impacts on SPEED, ROIC and Margin? • What would you do next as VP of SC? 43 43
Warning • Low hanging fruit is mostly picked • We are starting to find it requires more and more work to achieve smaller and smaller impacts • The harder we have to work, the more complex we make the operations – diminishing our impact further • We will have to think smart to be effective • Ideas? 44 44
Economies of Scale in Transport Class 70 LTL rates/lb from Green Bay to LA Shipment weight (lbs) 45 45
Questions? 46 46
Intro To GIS • Useful Tool: GIS • Geographic Information System – Links databases and maps – Useful for addressing questions like: • Where should we …? • How far away are…? • How are …. distributed?
GIS • Spatial data – Location – Stored in a shape file, geodatabase, … • Attribute data – Associated characteristics – Stored in DB • GIS systems typically maintain these separately and “join” them for display or analysis
Geographic Data • Raster model – Area covered by a grid of square cells – Image data is a Raster model, cells = pixels • Vector model – Features represented as • Points or nodes • Lines or arcs • Areas or polygons
Geographic Data • Key Properties of Geo Data – Projection: The method of translating locations on the 3 -D earth to (X, Y) coordinates on a map or screen – Scale: The ratio between distance on a map to the equivalent distance on earth – Accuracy • Positional: Is it in the right place? • Consistency: Is it correctly categorized? • Completeness: Is it all there? – Resolution: e. g. , pixel size
Layers • Different geo data maintained in different data layers with common location information, Smart Vector Layer: Raster. Map Digital Ortho photo: Different layers linked by Polygons describing land Lines defining the street combineslocations the visual common parcels network properties of a photograph with the positional accuracy of a map, in computer readable form.
Software Tools • • ESRI – Arc. View, Arc. GIS and others Map. Info – Map. Info and others MS – Map. Point (Tutorial, etc) Many others
Geocoding • Geocoding: – Placing data with location information on the map • Lat. Long • Street Address • Examples: – Geocode Stores by ZIP (Our. Company. xls) – Geocode Census Data by ZIP (US Census , Data. gov and private vendors sell demographic data) Bank Failures
Thematic Maps • Thematic Maps: – Display the attributes of geographic data • Examples: – Population • Sized Circles • Shaded Regions – Median Household Income • Shaded Regions
Geographic Queries • SQL-type query based on geographic information: – Eg. Average household income within 20 miles of Store # 1 • Map. Point is quite poor at this • The 2011 version can do crude selections • Example: – Household income within 20 miles of Store #1
Distance Calculations • As the crow flies – Using the Measure Distance Tool • Over the road – Using the Route Tool • • • Add Store #2 as Start Add Store #1 as End Get Directions Over the road distance Can adjust route options for speeds, hours of driving, rest stops, …
Territories • Given fixed centers develop territories based on proximity to those centers • Map. Point is very poor at this – Either you import them or you build them by hand… – Other (more expensive) packages do this automatically
Programming Tools • Map. Point includes software links (COM or Component Object Model) that allow you to integrate Map. Point into your applications (e. g. with VB, into Excel, Access, …) • We won’t go that far. • Interested: See About Programming with Map. Point • Embed within Excel, etc.


























































