Krify Articles
  Sign Up | Log In     Krify Home | Videos | Buzz | Free SMS | Matrimony | Answers | Web Hosting | Bulk SMS | Mail | News | Blogs


Music | Shopping | Forums | Freelance | Wall papers | Classifieds | Web director | Greetings | Bookmarks | Search | Games | Adman

Home Recent Articles Popular Articles Article of the day
Category list


Advertisements Ask Questions, Share Knowledge with Krify Answers
 
Posted by:  Raghu
 Article viewed:  2637  times



When To Use Simulation
Simulation has come a long way in the last 10 years, we now have affordable 3D simulation tools where the time taken to undertake the simulation is measured in weeks rather than months. Previously it was only cost effective to simulate large automated warehouses. These developments now mean that simulation can be justified on smaller traditional warehouses. This article looks at how simulation can be used, the data required and the benefits.



Definition

For this article simulation to means a model that steps though a period of time replicating the warehouse activity. There are many other models that are used for warehouse design, for example spreadsheets that calculate the number of moves per hour that a truck is capable of. Time based simulation looks at the interactions between different pieces of equipment as well as queuing and marshalling.

Simulation Types

Simulation can be used as a presentation device to show others how a new facility will look. Having animation with trucks moving makes the picture more realistic. This type of simulation can be set up quite quickly, as there is no need for actual demand data. If this is the case then it is important not to make decisions on the simulation outputs.

The next simplest simulation is to look at the mechanics of a single truck. Many fork lift companies have in-house tools to evaluate the performance of a machine in a defined role. These packages usually assume that there is always work waiting for the truck and progress is never interfered with. These tools are quick and easy to use and provide a valuable indication of the maximum performance possible.

The next stage is to set up a simulation with the assumed statistics of an operation. These statistics may include:

    * Average lorries received per hour

    * Average number of pallets per lorry

    * % of the pallets that are moved to each area

    * Average orders per day, average lines per order

By watching the simulation, it is now possible to detect congestion and bottlenecks. The number of pallets waiting to be put-away can be monitored. This makes sure the marshalling areas are not underestimated. Typical simulation outputs include equipment utilisation, average and maximum queue size, leadtime from order entry to despatch.

It is then possible to undertake a number of what if experiments:

    * What if pallets being despatched have a higher priority than put-away

    * What if trucks roam rather than work in specific areas

    * How many trucks are saved if there are 3 shifts rather than 2

The interactions in the warehouse become apparent, for example prioritising despatch decreases order lead time but may require more marshalling or more space for lorries to queue outside waiting to be tipped. Many results are counter-intuitive as a modern warehouse is complex with many interactions.

The GIGO (garbage in = garbage out) rule is especially true for simulation and if one of the statistics is wrong the results are suspect. To get most out of simulation it is important to concentrate on data collection and preparation. The preferred approach is to replay actual days with precise information about each transaction. For example a lorry arrived at 10:52am with 12 pallets of product X and 10 pallets of product Y. The simulation can then be set up with the warehouse management rules, which takes account of zoning.

It is possible to simulate a number of days, for example the peak inbound day, the peak outbound day or the peak order picking day.

Our approach for testing for growth is to limit the time available. For example if 25% growth is anticipated and there is an 8 hour window, the test is to complete the work in 6 hours.

The most detailed simulation is one where each individual product is assigned a pick location and actual orders are used. The exact picker activity can then be simulated and the best product layout determined. We have also taken the simulation to the store where the roll cages filled at the warehouse are unloaded and the shelf filling efficiency measured. In this way a compromise between warehouse efficiency and store efficiency was achieved.

When and when not to use Simulation

The obvious applications are where there is a high capital spend in automation. As well as ensuring a workable operation, simulation can avoid unnecessary expenditure. If there is to be a large number of operatives e.g. pickers, then reduction in head count will justify the use of a detailed simulation.

If there is insufficient data and the expectation is an accurate representation then simulation should be avoided as it leads to a false sense of security. If the equipment works in isolation with little interaction then a simple simulation will suffice.

Perhaps one of simulation’s most underused capability is as a tool to be used by operations to understand the interactions within their warehouse. Often the simulation is used to verify capital expenditure and then put on a shelf to gather dust, whereas operational management could use it to improve flows.

Using the simulation

    * Identify what decisions are to be made using the simulation, this determines the amount of data that needs to be collected.

    * Decide if 3D is required, it takes longer but is a much better tool for demonstrating both the simulation and the proposed facility.

    * Agree with operational staff the proposed flows and processes at an early stage.

    * Do not try and use it to decide between major design differences eg automation or traditional. This can usually be achieved using spreadsheet models at far lower cost.

    * Unless you have a dedicated resource to run the simulation and multiple applications, you would be better buying a service rather than a package

Disclaimer: The above article is responsible of the individual who post, krify.com does not hold responsible for any kind of disinformation. If you discover one or more of the krify.com pages direct you to messages that harass, abuse, have obscene, unlawful, defamatory, libellous, hateful, or otherwise objectionable content; or have spam, please inform to krify.com and that will be deleted as soon as possible.

Other interesting Article in the Category Matrixsim Articles
  Influence of Feelings In Life
  Quiz
  How To Live In A Simulation
  Building a Human Computer
  Desidarata
  Color of Life
  Plug Human Brain to Desktop PC
  Tips for Developing Value Metrics
  Virtual Reality

Write your Comment
Name:
Message:
 Verification code:    

Home Email to Friend
Google Pack

© Copyright - 2007 KRIFY SOFTWARE TECHNOLOGIES (P) LTD