Discrete Event Simulation and Its Role in Integrated Factory Modeling

MIKE JOLICOEUR: Hello, everyone. I'd like to welcome you to the discrete event simulation and its role in IFM or integrated factory modeling. My name is Mike Jolicoeur. I am a director of the Autodesk relationships for ProModel, which is a division of BigBear.ai.

So let's go through the class goals for the next 90 minutes. So just to make sure that you're in the right room, the class ID is IM501235, and here's the goals of what we're hoping to go through.

First, we want to make sure that you understand the benefits of discrete event simulation for manufacturing and AEC customers. We want to see how an ROI or Return on Investment can be developed from the data that is gathered from the simulation.

We want to be able to provide strong metrics to support a business case and also provide both 2D and 3D-animated process models utilizing AutoCAD, Inventor, and Forge with one button click.

So Autodesk tools provide a great foundation for building a digital twin and familiar tools that you've probably used for layout today. ProModel provides the ability to turn your CAD layout documentation into a powerful tool for throughput simulation, helping customers get the best return on investment for their process changes and in turn reducing costs of manufacturing of their products.

The Autodesk toolset provided in the product design and manufacturing suite does a great job of streamlining the layout process for space plane. The factory design utilities make it easy to transition to 3D and reap the benefits it provides, but the tool set lacks in regards to process planning.

That's where ProModel enters the picture. We provide the ability to test and evaluate design ideas and allow users to make the best decisions based on their criteria for success right inside the AutoCAD you use today for spatial layout.

In every project, large or small, there are three contributors to the design. Technical building equipment represents the connective tissue between the building and the production line. The building may represent either a new greenfield site or more commonly retrofitting to a brownfield site, or a combination of both of these. So hugely important in regards to successful installation of a new production area, this is not our topic really for today.

We are focused today on the production planning leg of the workflow. Spatial consideration is very important, but equally important, if not more important, is the performance of the process. Being that it is right sized for our production goals. That it delivers the correct amount of products for the lowest cost, and that it is agile in nature, allowing new products to be added without upsetting the balance of what is currently being produced.

A valuable tool for validation discrete event simulation is really nothing new. It's been around in one form or another since Henry Ford role Model Ts off the line. In fact, ProModel has been delivering PC-based simulation solutions for over 30 years to thousands of customers worldwide.

Discrete event simulation inside of AutoCAD, however, is new. The ProModel consulting team will tell you, over 95% of projects they delivered started with an AutoCAD layout. A lot of effort had to be expended to utilize a drawing in the ProModel environment, and for that matter in any other discrete event simulation tool that's out there.

So for that reason, ProModel and Autodesk teamed up to provide an easy-to-use efficient tool set that runs inside of the AutoCAD used today for layout. This new tool set helps build the foundation of an integrated factory model that is much more than a drawing that represents basically. The CAD model now becomes a powerful enterprise decision making tool.

So what exactly is process simulation? And what does it do? Well, the short version of it is it's a digital environment for experimentation, for validating ideas, or for improving your process either in existing production situations or completely new design ideas. It allows no risk experimentation with the process design ideas inside of the AutoCAD you use already today.

It's an easy to understand environment that ties your CAD layout to your process timing data and allows design idea iteration to get to the best answer. The data is the yin and yang of the process design.

If throughput can be achieved but it won't fit into the space allotted, that's no good. And if the layout fits great into the allotted space but it can't deliver the required products in the specified amount of time, it's also useless. So we have to have both to have really good and successful layout.

So ProModel provides time-based systems analysis. It mimics the operation of existing or future process for the purpose of understanding and improving that system. This also allows the ability to compress time. That is, we can simulate a year's worth of production in minutes and be able to see where the problems are.

Many times production issues in regards to throughput don't reveal themselves right away, and it can be challenging to find them without these tools as there are many interdependencies to track, which ProModel does for you.

So we also consider real world factors, one being variability and task times. We can account for variability and averages over multiple runs of a line to get the most accurate representation interdependent events. So we can figure out and identify where bottlenecks occur and figure out what is really causing them, not just what you think is causing them.

Resource availability, a major contributor to process problems. And it's nearly impossible to determine using spreadsheets. And it's easily understood in this environment. Also calendar factors, things like holidays, shift changes. Planned and unplanned downtime can be factored in, as well as maintenance timing studies and manpower needs, for example.

Broken down, there are three steps we follow for analysis. We first visualize the process. This is diagramming, modeling, and reporting what you see and the flow on the screen in process animation. We also then analyze the results.

So how do I increase service levels, determine resource requirements, or reduce costs, decide on the dials or levers that we're going to use or settings, in other words, and run scenarios so we can try different things and get different results? And then lastly, we optimize. This allows us to experiment, test, and explore in a risk-free environment. And it quantifies our successful KPIs and what we're looking forward to get out of the simulation.

So visualize. Let's go a little deeper on that. Within the AutoCAD environment, we can see the process run based on the criteria we set, conveyor system bottlenecks, process chokepoints, and spatial problems identified right in the AutoCAD environment.

Analyzing, so from here, we get configurable dashboards that allow us to measure against the key performance indicators that we deem most important, be that the takt time, part cost, or labor and resource utilization, or all of the above.

The chart tools are directly tied into the model so as design ideas are explored, the data dashboard updates as well. From the data provided by analysis, we can now experiment with the process via scenarios, pulling levers in the model to see how it behaves with different criteria being applied to that model. The dashboards allow us to then compare the scenarios against each other to help implement the best case that will offer the correct throughput at the best cost.

So the tools in the factory design utilities excel in helping customers avoid excessive cost and risk and changeover. That is making sure everything will physically fit into the space allotted and that the equipment interface is physically as well. This helps avoid change orders on installation which can drive project costs up substantially and also puts startup at risk.

What it does not show or does not do is allow planners to see how this plan change will affect the cost of the products being manufactured or the ability to confirm that the design will meet the required throughput to meet demand both in current state and in the future.

That's where ProModel comes in. We allow those that plan change, be it for new products, process improvement, or automation, or bringing acquired companies into the corporate fold, to fully understand if the process that is being designed will meet the production needs and be cost effective, either lowering the cost of manufacturing or allowing more manufacturing agility.

The combination of these tool sets allow customers to build out facilities digital twins that can be used over and over again to test ideas before committing any capital to the projects. The Autodesk ProModel solution reduces cost and risk not only for installation of the new process but operational risk for producing the product throughout its lifecycle.

We'll give you a brief overview of the user interface. A straightforward, easy to use, navigatable interface makes the tool easy for new users to master really fast. And experienced users also have the ability to apply macros, logic, and other advanced tools easily to further refine the simulation outcomes.

A tab is added to the AutoCAD ribbon containing all of the command structure necessary to create a simulation. Browser panels provide detailed information and interactive feedback in regards to the simulation being studied.

So let's see the products in action. Our first example is validation of an engine assembly line. Our criteria-- excuse me-- our criteria for success is listed here. We need an average throughput to be able to be met, but in addition, we need to be able to meet demand spikes as required.

Furthermore, we need to make sure that this fits into the space allotted and understand how the feeder products get to this line. This represents the intended flow of the line, timings and numbers based on past experience.

Note also that each of the times here are deterministic. That is, they have an absolute number. They don't have any variability in them. As humans are doing the work here for assembly, the chances of repeatability of timing is pretty low. That's why ProModel allows users to include process time variability into the model, increasing accuracy.

We find that when we build a model based on the existing flow that the customer gave us, we are typically within 5% of what the customer was seeing in the real world.

So as shown here, the AutoCAD layout can now be used to analyze our manufacturing design idea. The browser interface reflects the flow of the product through the line as seen here. We can see live how products will flow through the line. This will help designers complete tasks like rightsizing the line side AGV feed robots, as well as see where a potential bottleneck may be.

We can also gain a better understanding on how workers are traveling and help remove nonvalue added work and travel, as well as making sure we are properly staffed for the line.

Because we're using AutoCAD as the base and the CAD drawing is the scale, we can get distance travel for, part, material handling or workers, which are linked. Meaning that if we make a change to the drawing, the distance is update accordingly, reducing the chance for costly errors.

We can also apply scenarios which have different speeds and settings for the line and see what works best for our needs. Conveyor speeds, process speeds, manpower and staffing, all of these can be tested in the scenario manager. From here we get not only the dashboard output but the ability to show the simulation in 3D.

By clicking the Inventor button in the factory design utilities ribbon, the 3D representation is generated, making it easier to communicate our design intent with others. This can also be animated in 3D using ProModel's 3D simulation viewer, literally one button click. The easiest in the industry.

Using CAD visuals and the configurable output viewer, we can see where the problem lies. The station where heads are bolted on to the engine. Panning, zooming, and rotating the view while running are all supported in the viewer, allowing us to focus in as required.

Current design has the operator torqued in each bolt separately, which is time consuming not to mention error prone. We propose adding a multi spindle digital torque nut runner to the line, reducing cycle time and the chance for errors. However, this piece of equipment is around $259,000 US, which we will have to justify.

Scenarios can then be leveraged in the output viewer and make it simple for even nonCAD-oriented personnel to experiment with different configurations. The output viewer lets us clearly see problem areas represented in magenta for black processes, green for running, yellow for waiting, and blue for idle. Red representing downtime. The Throughput tab allows us to see that we were able to produce 12 more engines in two shifts.

So as we can see, we're able to validate space claim in 2D and 3D and validate that our line will produce more engines per shift, yielding 12 more engines per day based on the data that we had input into the simulation. It's projected that our profit on each of these engines produces around $984 per engine.

The estimated cost of adding a multi spindle nut driver to the line is around $245,000. Based on this, we can extrapolate an ROI of 20 days, giving confidence that the investment made in the equipment will pay back.

Moreover, you can reuse a simulation over and over again to test for different scenarios, process changes, or new processes based on this process. Or add more criteria to even further refine the process. Or to look at the process running over 16 months and see what the benefit over time is, including things like shifts, downtimes, and manpower constraints.

We can try other criteria as well based on the findings that the simulation provides, all in an environment that allows us to test these ideas before committing any capital.

Let's take a look at another example. A process retrofit that we've been trying to accomplish for over 30 years but always run into a roadblock when we reveal the price tag to make that change.

Sustainability is becoming a critical part of business decision making due to the cost of energy as well as being better to the planet. In this case, we'll use this to our advantage to sell this project to upper management and the financial controller.

In this example, we have a customer that is using a 1940s developed manufacturing process reliant on old technology machine tools that take a long time to change over and a natural gas fired heat treating process, which consumes a lot of natural gas and outputs a lot of carbon dioxide.

In addition, a substantial amount of space is required for the ovens, which, of course, needs to be climate controlled and electrically powered as well as maintained. Also in consideration, change over time.

When we originally commissioned this line, a 4-hour change over time for the analog bar feeding machines was acceptable as we had only six product variants. Natural gas was also inexpensive to provide fuel for the ovens that must run at 1,500 degrees Fahrenheit 24/7. Otherwise, we risk damage to the fire break linings of the ovens. Of course the financial controller says what he's said for the past 30 years. It works. The machines are long paid for. Let's just keep using it for another year.

Well, in today's production, we now have over 80 different variants of that art. So to offset change over time, we manufacture extras of each part, which becomes inventory. So this increases cost and promotes waste, both in space and old inventory that may or may not be used.

By using modern tool sets, we can use both spatial data and throughput data to our advantage. And in this case, factor in how energy costs can not only help us make the parts at higher quality and a lower cost but also reduce the carbon footprint, which may help improve the bottom line more through tax incentives for green initiatives. We can further reduce risk by using 3D layouts to ensure seamless installation and commission of the line.

Let's have a look at the tool in [AUDIO OUT]. Here we have a representation of current space. Analog screw machines, older three axis machines, and a gas fired heat treating process that requires its own building and also requires an automated guided vehicle to haul parts back and forth for construction process.

As earlier stated, these screw machines are not CNC. They're difficult to load, maintain, and takes a long time to set them up. Our design requires us to change only one part of the line, the beginning portion of the process, and eliminate the need for the gas fired heat treating ovens.

In the same drawing, we have a representation of the proposed process. We can tell ProModel to compare these two different scenarios. In this case, the current state being the baseline and our proposed future state. It's understood that a fairly-sizable investment needs to be made to modernize. So our goal is to show that the cost justified or the cost is justified for better line flow and reduced energy consumption.

We can start the simulation, and we'll see both systems running and running their parts based on the schedule that we provided. We're also tracking energy consumption on each machine and comparing the baseline against the proposed state.

In addition, we can see the simulation running in our 3D model representation or integrated factory model. The building that's shown in this view is a Revit model provided by our building design contractor, and is linked into the model.

If the building design changes, we'll see the update automatically in our 3D model as well as the 2D model. This prevents pollution issues when building or MEP changes are made, which, if they are not caught, can put the installation schedule at risk. We can also see that the current proposed models are displayed and run in 3D, making it better for communication of our design intent.

We also have a data table in the AutoCAD environment that's providing feedback on the line performance based on our criteria. We can see that a substantial energy savings is realized, and a dramatic reduction in carbon footprint is also realized. We also get the data dashboard that tells us all we need to know in regards to cost throughput and manpower needs.

So from here, we can see the summary of all of the operations that are happening, and we also have the ability to look at each of the operations as compared to each other.

So let's recap the results. With a sustainable, friendly workflow, not only do we consume 65% less energy, which reduces costs and has positive tax opportunities, we drive down the cost of the parts and produce more parts with less, and impressive 73% reduction in energy costs per part, which in turn reduces the production costs of each of those parts. This doesn't include the benefit of unneeded inventory, but that could be measured as well to further build out our case.

Also, 69% less carbon dioxide getting released into the environment to the tune of nearly 500,000 pounds of carbon dioxide reduced each year. With a sustainable, friendly workflow, the extrapolated benefit would pay for two of the new machines in the first year. Our outcome is a well supported, well documented reason to change that finance people can't argue with.

Last example is a logistics study. So this customer built in four lines into their new distribution center, and we're currently using only three of them. When they turned on the fourth line because of an increase of demand, the outcome wasn't quite what they expected. They were having trouble determining where the problem was, and they were considering adding more equipment or expanding space based on past experience.

So we worked with the customer, and we built out a simulation of the process, both the as is state and the new state running the fourth line. We were able to determine from the data and the flow the new line was feeding the system at two faster rate, disrupting the flow of the existing system. In short, adding the fourth line did the exact opposite of what they expected. It slowed the system down. It raised costs due to overtime being needed to keep up.

From the data and by watching the simulation, we can see where the problem lies, and we have a good idea on where to start for correcting this problem. According to the simulation, the proposed change gets rid of congestion, allows the entire line to flow efficiently within a standard 40-hour workday, or work week, I should say. So as we can see, the congestion is now eased in here, and we have a smoother running line.

Let's take a look at it running actually, and you should be able to easily see the difference. So the change in the design of the pellet feed into the warehouse was a simple, low cost change as compared to the other options of building expansion and equipment additions, which in reality probably wouldn't have solved it.

As we run the current against the proposed, it becomes obvious where the problem is. If you look at the left side of the screen, you can see that the conveyor is backing up in the Azure simulation. The proposed change, which was simply a change to the conveyor speeds and smarter feeding into the system, alleviated the backup and it solved the problem.

By applying the entire solution, cost savings and design confidence can be assured not only in the installation phase but also validation that the design will meet production output goals with financial information to back it up. In all cases, we proved our ideas with clear, actionable data that can be used over and over again for testing new scenarios of production.

By applying the entire solution, cost savings and design confidence can be assured in not only the installation phase but also validation that the design will meet production output goals with financial information to support it. We have complete design confidence.

So would you like to know more about the ProModel solution and how it interfaces with the AutoCAD tools that you're using already today? Well, come over and see us over at booth 220. We have consulting staff available to answer further questions.

We are also going to be having special show pricing for ProModel, and that's where people that show up at the booth with a coupon. And we can explain more of how that program works to the people that come. And reseller partners are also available to provide software and training. In fact, some of these reseller partners may already be the partner that you're working with.

I'd like to take-- I would like to thank you for your time and attention. Hopefully, you found this to be a useful set of tools and a useful workflow to help you and your company be able to save money and make things a lot more efficient.