Simulation

Simulation can be a separate service offered by IPACT.  More often, it is an integral part of a project.  It can be one of a number of different types.  But no matter the “flavor”, simulation is a strategic and vital part of IPACT’s automation methodology, and a major component in distinguishing us from our competitors.  Everyone says that they do simulation, however it pales in comparison to what IPACT delivers!

What is Simulation?

Broadly defined, when simulation is part of a larger automation project, it is a piece of automation that acts like the plant and its existing automation.  The new automation under development in the project can be connected to the simulation, and then be tested under conditions that will closely approximate what will really happen when the new automation is installed in the plant.  (There is another type of simulation, which we call a process simulation, that has a different goal and is discussed below.)

How Does Simulation Help?

Fundamentally, simulation helps both IPACT and its customers by insuring that the new automation being supplied will perform as expected.  There’s nothing like putting the system into the (virtual) plant environment and watching it run to test its performance.  Specifically:

• Simulation provides an environment to run the Factory Acceptance Test (FAT) against.  In some sense, it makes a good FAT possible.
• Simulation allows you to “extreme test” the system in ways you cannot on the plant floor.  You can break things, run them crazy fast or slow, make tanks overflow, pipes burst, and conveyors seize up – all with no lost production, mess, or repair bills.
• Simulation provides a realistic environment in which plant operators and technicians can be trained.  Our customers have taken advantage of this many times, usually from their plant location (remotely), with the simulation running at our office.
• A simulation provides an environment within which anomalies seen in the field can be reproduced and tracked down much more easily than on the “live” system.  Click here for an example.
• Finally and most importantly, simulation is the key to a system that starts up smoothly.  Poor startups are the “dirty little secret” of manufacturing automation, and we continue to be amazed at what kind of chaos customers put up with from other suppliers.  What does it cost for your line to be down or crippled for a day?  A week?  A month?  At IPACT we pride ourselves in our fast, smooth, “boring” startups.

Types of Simulation

I/O Simulation

This type simulates the response of the plant from the perspective of the new automation system’s digital and analog inputs and outputs.  These would typically be outputs like motor starters, valve open/closed commands, and speed references, and inputs like limit switches, tank levels, and hopper weights.  The simulation “watches” the digital and analog outputs from the new system, models the response of the equipment, and sends back the digital and analog inputs to the new system.  Besides modeling the response of devices like valves and motors, the simulation must also model movements of materials, heating and cooling of product, etc.  This is the most common type of simulation and is done for the great majority of IPACT-supplied new automation solutions.

MES Transaction Simulation

This type simulates the data transactions that occur between the new system and other automation systems in the plant.  These other automation systems are usually business systems (e.g. SAP), but can also be quality systems, laboratory information systems, etc.  Sometimes, if the new automation system does not communicate with plant I/O, this is the only type of simulation that is necessary.  Other times, both an I/O simulation and a MES transaction simulation are required, as the new automation system will have connections both to the plant I/O and to other computer systems.  IPACT has a simulation product that allows us to quickly implement MES transaction simulations.

Process Simulation

This is a different type of simulation than we have discussed so far, and has a different goal.  The goal of I/O and data transaction simulations is to provide a virtual plant environment for a new automation system to work in, primarily for the purposes of testing.  A process simulation is not associated with a new automation system – it is self-contained and its goal is to model how a plant will behave over time.  IPACT has done numerous process simulations for our customers, and an example here will provide the best explanation.

IPACT had a pet food client that was about to make a major expansion.  The centerpiece of the expansion was a matrix of 24 large storage bins.  These bins were fed incoming material from four extruders through a handful of shared paths.  Six packaging lines pulled material out of these bins for blending and packaging, also through a number of shared paths.  Often one packaging recipe would call for three or four different components, requiring paths to several bins for that one packaging line. 

Our client was concerned about the shared input and output paths.  What would be the effect on production of both extruders and packaging lines having to contend for paths to/from the bins, especially as a production week evolved?  This question is virtually impossible to answer from a theoretical point of view, and most commercially available simulation packages do not cope well with this kind of problem.

Before construction began, IPACT worked with our client to construct a simulation of the extruders, bins, and packagers that graphically displayed the process and ran 60 times as fast as actual production.  This meant a whole shift would be simulated in eight minutes, and the entire five day production week in two hours.  The client’s production schedulers would give the simulation a set of extruder schedules and packaging schedules for a production week, and then sit back to watch what happened.

Indeed, the client found that for many types of schedules, the system would eventually become “gridlocked”, with all bins full, thereby blocking extruders, but no packagers able to reach all their required components from the bins.  Watching the simulation suggested to the client what the solution might be, and the shared paths were modified and expanded in the simulation.  This solved the problem and kept production flowing smoothly.  The design was changed for this new equipment configuration, and the system performed well once it came on-line.

Finding this problem in simulation ahead of time saved our client an immense amount of lost production time and therefore money.  This problem would have taken weeks and huge $$ to retrofit in the field if it would have not been discovered until the system was in production.