M AT E R I A L

HANDLING
S I M U L AT I O N

W H I T E PA P E R
TABLE OF CONTENTS

01 Why Use Simulation in Material Handling? 02

02 Key Material Handling Challenges 05

03 Modeling How Your Organization Works: Simulation Versus Linear Programming and Spreadsheets 10

04 Walmart’s Alphabot: Designing a Material Handling System with Simulation  13

05 Improving Reliability and Profitability for a Large European Steel Manufacturer 18

06 How to Get Started with Simulation 23

Conclusion 25

Additional Resources 26

01
WH Y US E
SIM U L AT I ON
IN MATERI A L
HAND L I N G ?
 WHY USE SIMULATION IN MATERIAL HANDLING?

WHY USE SIMULATION IN MATERIAL HANDLING?

Material handling in facilities as diverse as steel mills and retail fulfillment
centers is benefiting from simulation-based analysis and design. Detailed process
insights and forecasting advance decision support beyond spreadsheet-based
and top-down analysis. When working to resolve material handling problems
such as debottlenecking, balancing storage constraints, and optimizing
production plans, simulation modeling provides advantages not offered by
traditional techniques.

Simulation modeling is effective for complex and large-scale systems where

dependencies might be unclear. It solves real-world problems safely and
efficiently in a virtual environment and, thanks to detailed statistical and visual
outputs, enables analysis which is easily verified, communicated, and understood.
For material handling operations, simulation modeling gives detailed and clear
insights for well-informed decision making.

03
 WHY USE SIMULATION IN MATERIAL HANDLING?

Key uses of simulation for material handling include: Overall, organizations that make use of simulation modeling are more informed
and agile when it comes to making changes to how they work and are better
• Facility design and capacity planning to prepared and resilient in the face of disruption.
improve ROI or throughput
If you or your organization has yet to try simulation-based material handling
• Process and scheduling optimization to increase analysis, it is easy to get started. Learning is free with AnyLogic PLE and there
production and service levels are many resources to help you along the way. Read on to discover more
about material handling simulation and how organizations use it to solve
• Equipment usage analysis for calculating ROI business challenges.

• Bottleneck analysis when trying to mitigate and

balance constraints

• Inventory planning when seeking lower waste and

increased productivity

• Determining staff and equipment requirements to meet

production flow demands

• Performance metric analysis

Simulation modeling is especially useful when making speculative changes on a

real system is impossible or impractical because of costs or time.

The ability to analyze a model as it runs sets simulation modeling apart from
other methods, such as those using Excel or linear programming. The freedom
to inspect processes and interact with a simulation model in action helps quickly
build both understanding and trust.

04
K E Y M AT E R I A L
HANDLING
CHALLENGES
 KEY MATERIAL HANDLING CHALLENGES

KEY MATERIAL HANDLING CHALLENGES

Material handling operations present many challenges that benefit from to low machine availability, delivery processing times, and storage capacity.
simulation modeling and analysis. Balancing end-to-end material flow requirements is important for controlling
costs and increasing profit.

Failing to capture the details necessary to accurately analyze and forecast

production or warehouse operations can lead to incorrect insights and
decision making. Some problems with traditional spreadsheet-based
throughput analysis include:

• Fixed data snapshots can make troubleshooting difficult. For

example, daily stock level snapshots may make bottlenecks
caused by storage and retrieval inefficiencies hard to find.

• Numbers and charts do not allow processes to be visually

verified. Analysts and managers with experience of material
flows in a facility will need time to apply their operational
knowledge to static tables of numbers and charts.

Reliable information about current and predicted performance makes it easier to

optimize processes and implement changes. Simulation makes good throughput
analysis and optimization possible because it can authentically represent a
THROUGHPUT ANALYSIS system’s underlying details, dynamics, and dependencies. Beyond static tables
and charts, simulation provides dynamic views and process visualizations based
The throughput of a facility depends on economic, environmental, on the capture of a facility’s processes and workings.
and equipment-related factors which are often complex, dynamic, and
interconnected. Examples include staffing levels, production bottlenecks due

06
SCHEDULING AND SEQUENCING

Production and storage facilities involve elements that are subject to external
factors such as suppliers, purchasers, and transport. Even inside a facility,
accurate forecasting and optimization can become complex. Challenges
include maintenance requirements, seasonal demand fluctuations, customized
production, and finding optimal order picking and processing algorithms.

Manual or constraint-based scheduling based on spreadsheets or mathematical

models can miss important details or relationships, especially in large systems.
Difficulties can be found in the following common areas:

• Order driven manufacturing means coordinating many dynamic

operations and processes. In a JIT, or lean manufacturing
environment, meeting scheduling KPI can quickly become too
complex or time-consuming. Simulation makes clear the behavior
of a system and does not need a complete mathematical model
because it accounts for each element and how they interact.

• Machine maintenance scheduling may not account for

the state and wear of machinery, leading to unnecessary or
unexpected downtime. With the simulation of machinery,
the processes governing usage and maintenance can be
captured so that adaptive and more efficient maintenance
scheduling can be applied.

• The spatial requirements of people and machinery are difficult Simulation-based scheduling accounts for time-based and causal dependencies
to include when planning changes using traditional scheduling and can accommodate large scale systems made up of numerous entities. It
methods. With simulation, the dimensions of machinery and shows the behavior of a system and the possibility to derive a schedule by
spatial requirements for staff can be made part of the model. modeling each of a system’s components and how they interact. When a system

07
 KEY MATERIAL HANDLING CHALLENGES

is non-linear and dependencies are unclear, simulation-based scheduling is a them to understand the likely impact of new equipment and determine how best
powerful tool for analysis and testing. to integrate it into facility workflows.

PLANNING FOR NEW EQUIPMENT

OPERATIONS AND POLICY ANALYSIS
When introducing new equipment such as automated guided vehicles (AGV),
conveyors, cranes, heavy machinery, and automated storage and retrieval Experimentation with live operational processes and policies can be disruptive
systems (ASRS), it can be difficult to accurately predict their wider impact and and counterproductive. Safety and cost considerations can also limit real-
return on investment. But, without a good understanding of the effects the world testing. The alternative is to plan and test a system using a model.
introduction of new equipment will have, it becomes impossible to optimally However, while spreadsheets and linear programming offer some virtual testing
configure a facility. capabilities, they cannot properly account for stochastic events such as machine
failure or time-based interdependencies like customized orders that involve
Simulation offers an alternative method of analysis when traditional methods, multiple production lines. For cases like those, simulation can provide a solution.
such as those based on spreadsheets, are unable to scale sufficiently or cannot
represent a system’s behavior accurately. Examples include: A simulation model of a factory, port, or warehouse will represent each process-
element, how these elements operate, and the rules that determine their
• Large scale facilities with complex interactions such as at interaction. This capture of a facility’s processes enables experimentation
ports, on production lines, or in warehouses. and analysis in a risk-free environment. Policy experiments can be conducted
using the model and, when they look promising, they can be refined and
• Accounting for the impact of new equipment on a project’s tested against forecasts and historical data. The result of using simulation
overall profitability. to analyze and develop policies is more agile decision making and greater
implementation success.
• Determining the environmental impact of workflow changes
related to new equipment. Simulation modeling is used when:

With simulation, a digital replica of your current systems can be made so that • Carrying out Material Requirements Planning (MRP)
extensive testing and planning is possible without affecting ongoing operations.
Decision-makers and managers can use simulation as a ‘what-if’ tool that allows • Debottlenecking storage and retrieval systems

08
 KEY MATERIAL HANDLING CHALLENGES

• Evaluating alternative Bills of Materials (BOM) Before committing to changes, their consequences should be predicted,
analyzed, and understood. Beyond spreadsheet-based and linear modeling
• Applying a theory of constraints approach to techniques, simulation can provide a method that accounts for stochasticity and
manufacturing operations complexity as well as easing verification with 2D and 3D visuals.

• Testing machine learning developed maintenance policies

09
M O D E L I N G H OW
YO U R O RG A N I Z AT I O N
WO R K S : S I M U L AT I O N
VERSUS LINEAR
P RO G R A M M I N G A N D
SPREADSHEETS
 MODELING HOW YOUR ORGANIZATION WORKS: SIMULATION VERSUS
LINEAR PROGRAMMING AND SPREADSHEETS

MODELING HOW YOUR ORGANIZATION WORKS:

SIMULATION VERSUS LINEAR PROGRAMMING AND
SPREADSHEETS
The importance of spreadsheet models in business planning is clear from
their widespread use.

However, despite their importance, more than 90% of spreadsheets contain

errors and around 50% of spreadsheet models in large organizations contain
material defects (EuSpRiG). As a result, companies can put themselves at
unnecessary risk by acting on the outputs of these spreadsheets.

A more reliable method of modeling that helps inform decision making is linear
programming and it is widely used to help optimize planning and production. It is
a more reliable method because the work required to model a system with linear
programming helps overcome many of the problems spreadsheets have with
their lack of verification.
Agent-based, Discrete event, and System Dynamics modeling
To model a system with linear programming requires knowledge of its functional
dependencies from input to output. Often these dependencies are not clear
because they are causal, time-related, or just because the model is large. In the come from it being a verifiable method of modeling that can deal with large
end, these system models become hard to manage, difficult to communicate, and complex systems, time dependencies, and stochasticity.
generate low-value insight.
Simulation models work by capturing the characteristics and behavior of a
For complex material handling problems, simulation offers an alternative method system’s components. Then, by observing how the components interact, it
of analysis to spreadsheets and linear programming. The strengths of simulation is possible to see the behavior of the whole system. A simulation can output

11
 MODELING HOW YOUR ORGANIZATION WORKS: SIMULATION VERSUS
LINEAR PROGRAMMING AND SPREADSHEETS

tables of data and charts as well as animated 2D and 3D visualizations.

With the possibility to inspect the logic of each element and to verify
behavior with visualizations, a simulation model can easily be checked and
communicated at multiple levels.

Simulations developed in AnyLogic can combine the three main methods of

simulation modeling: agent-based, discrete event, and system dynamics. The
combination of these methods means a simulation model can accurately reflect
the elements of a system, their processes, and how they interact.

With simulation, you have a method of analysis for material handling

challenges that involve dynamic and stochastic processes in large-scale complex
systems. Examples include developing an autonomous AGV order fulfillment
system for Walmart and improving returns on investment at a larger European
steel manufacturer, as can be seen in the following case studies.

12
 WALMART’S ALPHABOT: DESIGNING A MATERIAL HANDLING SYSTEM WITH SIMULATION

WA L M A RT ’ S
A L P H A B OT
DESIGNING A MATERIAL HANDLING
SYSTEM WITH SIMULATION
 WALMART’S ALPHABOT: DESIGNING A MATERIAL HANDLING SYSTEM WITH SIMULATION

PROBLEM

Walmart, the world’s largest retailer by revenue, was looking for an automation
technology for its fast-growing online grocery business that would help complete
orders faster and at a lower cost. They chose to evaluate Alert Innovation’s
Goods-to-Person (GTP) concept, Alphabot® – a system that can automate online
grocery at the store level by using autonomous mobile robots that are capable
of operating in all three dimensions of a multilevel storage structure. Alphabot
robots are self-driving vehicles that can gather items from the ambient, chilled,
and frozen temperature zones of a high-density storage system and bring them
to associates who are picking items to fulfill customer orders. Alert Innovation
invented Alphabot® to make in-store fulfillment of online orders faster
and more efficient.

The retailer wanted to evaluate the feasibility of the Alphabot® concept, and
its suitability for Walmart, before making a financial commitment towards
product development. Alert Innovation had already made some static
spreadsheet calculations for the project, however, both Alert Innovation and
Walmart agreed that the spreadsheets could not be relied upon due to system
complexity and variability in demand and execution. As a result, and before
making a financial commitment to deployment, MOSIMTEC, a simulation
consulting firm, was tasked with designing a material handling simulation model
for an independent technology feasibility assessment. The goal of this initial
modeling assessment was to:

• Understand the throughput and turnaround time capabilities

of the Alphabot® system during peak customer demand periods
and assess service levels.

• Understand any potential performance weaknesses of the

Alphabot® system with realistic demand data and variability.

14
 WALMART’S ALPHABOT: DESIGNING A MATERIAL HANDLING SYSTEM WITH SIMULATION

• Identify the best configuration for the racking structure, Furthermore, Alphabot®’s programmers needed to develop numerous control
automated bots, and picking workstations required for algorithms and AnyLogic’s Java capabilities meant algorithm ideas could be
different store profiles. tested quickly and easily.

• Compare the capital expenditure requirements predicted by Stage 1: Material handling simulation model animation including
the model against the business case spreadsheets. AGV simulation

• Use model throughput and turn-around-time results to Walmart’s initial goal was to make a go/no-go decision about the Alphabot®
compare the proposed performance of Alert Innovation’s project. To support this decision making, MOSITMEC learned the system,
Alphabot® with other industry-leading Automated
Storage and Retrieval Systems (ASRS) and Goods-to-Person
(GTP) technologies.

Material handling simulation would not only help understand the true cost
of rolling out the system in the real world, but also identify store-specific
requirements for the potential deployment of Alphabot® across Walmart’s other
stores in the future.

SOLUTION

To model Alphabot®’s behavior and operations in a computer simulation, with

real-world complexity and variability, MOSIMTEC chose AnyLogic simulation
software. MOSIMTEC used AnyLogic’s abilities to dynamically build facility
layouts from input data to help cut model development time and enable faster
evaluation of multiple Alphabot® configurations.

15
 WALMART’S ALPHABOT: DESIGNING A MATERIAL HANDLING SYSTEM WITH SIMULATION

designed initial bot control algorithms, built a material handling simulation • More bot-specific movement logic: bot paths, reservations,
model, conducted an analysis, and presented its findings to Walmart’s leadership. collision avoidance logic, and movement profiles, using AnyLogic
agent-based modeling and Java capabilities.
In the final delivered model, Walmart managers could specify different inputs
like the number of bots, their length, width, acceleration, and speed in different • Separate temperature zone storage system logic for shelf-
areas. Physical racking configurations such as the number of aisles, levels, stable, refrigerated, and frozen items.
the space between levels, and the number of workstation tiers, along with
other physical components of the system, were also configurable. Control- • Inventory tracking functionality within the racks to track every
logic parameters, including those for work-assignment and for setting various item in the system.
thresholds, were also exposed and made available.
• Bot sequencing logic to help simulate robots sorting items.
Stage 2: Material handling simulation model animation including
AGV simulation With the updated material handling simulation, engineers were able to
validate the original system design assumptions and provide feedback
Model input and output statistics were connected to an Excel front-end so to Alphabot® product development teams. The Alphabot system model
that users could easily configure and run the model. At this concept evaluation including AGV simulation
stage, MOSIMTEC also incorporated a 3D model animation that was scalable
based on the layout defined by the user within Excel. The output results in Excel
included a summary report with key metrics, log files, scenario comparisons,
charts, and graphs. OUTCOME

After completing the independent analysis of system capabilities, MOSIMTEC The initial material handling design model helped make the business case for
transitioned the AnyLogic material handling design model to Alert Innovation for deploying Alphabot® more widely by showing the equipment requirements
long-term use in fine-tuning software control algorithms for eventual production needed to meet various target turn-around-times. The simulation model also
deployment. Based on the outcome of Phase 1, Walmart proceeded with its quantified system performance limits when subjected to unconstrained demand
Alphabot® system development. Alert Innovation engineers used the Phase and showed that Alphabot® would be able to pick 95% of real-world online
1 model as the foundation for Phase 2 which increased the level of detail and grocery orders in less than eight minutes, with an average picking time of
testing of control algorithms to simulate different system design alternatives. under five minutes.
Model enhancements included:

16
 WALMART’S ALPHABOT: DESIGNING A MATERIAL HANDLING SYSTEM WITH SIMULATION

As the project progressed, the initial model was updated and expanded to
better understand the impact of various detailed design alternatives. The model
helped Alert Innovation determine which design alternatives would result in the
greatest ROI and how the system can be configured for future deployments.

Walmart and Alert Innovation launched a proof-of-concept pilot of Alphabot® at

a Walmart supercenter in Salem, New Hampshire in March 2019.

Watch the video of Amy Brown Greer, Dr. Christian Hammel, and John

Lert presenting this case study at The AnyLogic Conference.

17
 IMPROVING RELIABILITY AND PROFITABILITY FOR A LARGE EUROPEAN STEEL MANUFACTURER

G O L D R AT T
R E S E A RC H L A B S
IMPROVING RELIABILITY AND
PROFITABILITY FOR A LARGE
EUROPEAN STEEL MANUFACTURER
 IMPROVING RELIABILITY AND PROFITABILITY FOR A LARGE EUROPEAN STEEL MANUFACTURER

PROBLEM

One of the largest and oldest European steel manufacturers was trying to use
lean manufacturing tools to improve service levels and returns on investment.
Customers and investors were demanding much higher performance, but efforts
were falling short and significant challenges remained. Their initiatives ended
up in repeated debottlenecking and a stable system seemed unobtainable. The
company called on the assistance of Theory-of-Constraints specialists Goldratt
Research Labs. Together, they decided to transform and optimize the company’s
supply chain, taking into account its high degree of fragility to external changes,
low profitability, and low ROI.

Complexity made supply chain changes difficult and it was risky to try new
policies and rules because outcomes were unpredictable. The company’s static
decision support tools, such as their enterprise resource planning system and
various spreadsheets, did not help because of difficulties working with the
complex operational and business environment.

To fully consider all the interdependencies, constraints, dynamics, and variability

in the system, the system analysts and engineers decided to use simulation
modeling. The engineers could capture the complex details of the manufacturing
supply chain, determine the causes of performance gaps, and test possible
solutions in a digital environment. Simulation modeling would help management
make faster, better decisions and predictions, and turn these predictions into
reliable manufacturing optimization commitments for investors and customers.

19
 IMPROVING RELIABILITY AND PROFITABILITY FOR A LARGE EUROPEAN STEEL MANUFACTURER

SOLUTION With the coil locations and movements known, the model helped identify
efficient options for congestion reduction based on throughput, in-process
Several models were built that together could represent the whole integrated inventory, and cost impact. Operational changes, including increasing hot
steel production process. The developers took advantage of AnyLogic’s strip mill output and coil width doubling, were tested for better production
multimethod modeling capabilities which made it possible to reflect the planning and control.
components, processes, and interdependencies in the system. The models could
show outcomes for single scenarios, conduct sensitivity analysis for various The model also helped find the best way to upgrade and optimize the wagon
parameters, and provide direct scenario comparison. fleet. In the future, management plans to use the model for testing operational
rules that allow full automation of the hot coil management process.
AnyLogic can work with external data and self-configure models and, in this case,
ingested more than 70 Excel worksheets’ worth of data from the enterprise
resource planning and enterprise management systems. AnyLogic ability to
create self-configurable models from external data helped cut development
time significantly.

The manufacturing process simulation models have detailed 2D and 3D

animation. The AnyLogic platform provides easy-to-use interfaces and the ability
to export models as standalone applications to run any computer, without the
need for special software.

Furthermore, mobile work and collaboration were facilitated by the browser-

based AnyLogic Cloud.

Hot Coil Finishing Simulation Model

The hot coil finishing area was significantly congested and there were constant
coil movements related to coil retrieval. Even the car park was filled with coils.
The need for change was acute.

A model was created that could show the location of all the coils in the system.
For each coil, the type, the destination, and the storage location were specified.

20
 IMPROVING RELIABILITY AND PROFITABILITY FOR A LARGE EUROPEAN STEEL MANUFACTURER

Steel and Slab Simulation Model

The steel and slab production simulation model made it possible to:

Steel production operations and their scheduling are both very complex and it • Capture and digitize the scheduling and operating rules that
is very difficult to predict the impact of changes in product mixes or operational different plant operators used. These rules had never been
rules. To test them and improve efficiency, a specific model was needed. verbalized before and could eventually be lost. Now personnel
can learn from shared experience.
The model that was developed captured all the elements of the manufacturing
system. In the 2D animation, clicking a crane or product shows its operational • Develop scheduling and operating rules to avoid
status. At the heart of the model is the logic for production planning and the bottlenecks. For this, different speed restrictions and
scheduling of operations. The required product mix and other parameters can be alternative prioritization rules for different products were
set and altered as needed. tested and analyzed.

21
 IMPROVING RELIABILITY AND PROFITABILITY FOR A LARGE EUROPEAN STEEL MANUFACTURER

• Forecast the operational and financial impact of producing

more complex and higher-margin products, which could
negatively affect workflows.

OUTCOME

Using the models, the causes of congestion and inefficiency were found.
The models provided a safe, low cost, and quick way to test the impact of
operational and financial changes. The models were also used to validate
each other’s results.

The models are now used on a weekly, monthly, and annual basis to analyze
workflows and inform critical decision making. The results provide reliable
forecasts and enable commitments to be met. Project presentation by Dr. Alan
Barnard and Jaco-Ben Vosloo from Goldratt Research Labs.

Other related case studies:

Cardinal Health – Optimizing Warehouse Operations

Kuehne+Nagel – Warehouse Picking Algorithm Evaluation

GE – Real-Time Manufacturing Decision Support

22
H OW TO G E T STA RT E D
W I T H S I M U L AT I O N
 HOW TO GET STARTED WITH SIMULATION

HOW TO GET STARTED WITH SIMULATION

Simulation modeling is a powerful tool for analyzing and developing complex • Solar Panel Production
material handling systems that can provide unique strategic and operational
insight and decision support. But how do you get started? • Cranes

If your organization has not used simulation modeling before, it can be difficult • Network ports
to know where to begin. AnyLogic has many resources to help ease development
and understand the important factors that need to be considered. To get • Restricting areas for transporters
started quickly, it is an advantage to have a background in areas related to
systems analysis and integration, industrial engineering, or manufacturing The AnyLogic team and our expert partners around the world are experienced
process engineering. For advanced features, it will also help to have some in the application of material handling simulation. Find your local AnyLogic
knowledge of programming. specialist, or contact us directly to find out how simulation can be
applied to your challenges.
To get an idea of how simulation modeling is being used by companies to solve
material handling challenges, there are case studies and resources available in
the AnyLogic Material Handling Library Pack.

For those who wish to start by trying the software, AnyLogic PLE offers a free
way to begin. Included in the application are tutorials with accompanying models.
There are also how-to videos, webinars, and material handling specific in-
depth tutorial blogs:

• Conveyors

• AGVs and transporters

24
 CONCLUSION

CONCLUSION
Simulation modeling is an advanced method for analyzing and developing
material handling systems. It offers an alternative to the traditional methods
of spreadsheet-based analysis or linear programming and can be superior when
developing solutions for complex manufacturing and warehouse processes.

Material handling business challenges that can benefit from simulation include
throughput analysis, production scheduling, impact assessment for new
equipment, and the analysis and development of operations policies.

AnyLogic is multi-method simulation software that combines the three main

modeling paradigms: discrete event, system dynamics, and agent-based.
Combining these methods allows systems to be modeled in detail and
without compromise.

Specific modeling libraries in AnyLogic also assist modeling. The Material

Handling Library, Process Modeling Library, Rail Library, Road Library, and
Fluid Library combine to help capture the details of a system. The combination
of multi-method modeling and these libraries is a platform that can capture
material handling systems in detail to give unique insights and decision support.

When traditional modeling methods are not sufficient, simulation modeling

provides a powerful alternative.

25
ADDITIONAL White Papers

Case Studies
RESOURCES
Videos

Books

Download AnyLogic Simulation Software

Seminars and Training

26
CONTAC TS
TH E AN YLOG IC COM PANY

info@anylogic.com
Contact your local office >>

27