8. Smart Manufacturing - The Factory of The Future

by Dale Whitfield, October 2014

What can we expect to see in the Factory of The Future?

The Factory of the Future will be built around an Internet of Things (IoT). The concept and examples have been around for a while under the guise of Machine-to-Machine (M2M) communications and various other headings. Ultimately, everything is connected. Machines are connected to other machines.

The German company Fraunhofer "Europe's largest application-oriented research organization" says the following:
"The "Integrated Industry" sector has, thus, been in the focus of many research activities at Fraunhofer for some time. Cyber-physical systems link the virtual world to objects of the real world. This creates the "Internet of things" where real products and production processes are linked to the Internet. This accelerates processes and helps to better utilize resources. However, what is decisive for the success of the integrated industry, is the human factor."

A new way to look at the embedded systems of tomorrow is encapsulated in the concept of Cyber-physical systems (CPS)

The US National Science Foundation (NSF) stated:
"Cyber-physical systems (CPS) are engineered systems that are built from, and depend upon, the seamless integration of computational algorithms and physical components. Advances in CPS will enable capability, adaptability, scalability, resiliency, safety, security, and usability that will far exceed the simple embedded systems of today."

Standards for inter-operability that will facilitate machine interactions are in the process of being developed.

How do we prepare for this with our products?

Some initial thoughts:

• Interconnectedness of machines
• Machine awareness of its role and its effectiveness in the production process
• Machine Identity - what role and functions it is able to perform
• Auto configuration of peripheral devices
• Cabinet free machine designs
• Move the HMI / UI off the production floor
• IT barriers will come down and we need to be ready
• Industry 4.0 - German Government strategy to promote computerization of manufacturing industry. The intelligent / Smart Factory. Connects into the IoT. Reality in 10+ yrs.

To be continued...

7. Buzzwords! The Internet of Things, The Cloud

by Dale Whitfield, November 2013

What is The Internet of Things?

Many devices today are becoming connected in the sense that they can communicate via the Internet. Given that they are devices and not people on computers, phones or tablets, they are just things. Technically, if a device has an IP Address, it is able to be connected via the Internet.

How does this help us?

There isn't much to do, it all falls into place and takes care of itself.

The IPAV System deploys a number of devices around a production line at what are called points-of-interest. Typically that's one point at each significant machine on the production line.

Since the data flowing between these points-of-interest needs to be in real-time and secure, these devices are not individually addressable, in the IP address sense. There is also no real justification for making them addressable. They could be, but its more of a hindrance than of use.

The Globe device is Internet addressable and is joined to that network of devices on the production floor. It's physically located away from the production area and may be connected into the plant's Local Area Network.

Note that the data moved around always maintains individual machine data in its structure. Even though there is only one connected device, this represents all the devices at their points-of-interest. Now the analysis and processing of this data can be carried out, away from the production process.

Once the Globe device is on the LAN, it may be made available into the wider area known as the Internet. Data is now able to be securely moved from the plant to other plants, head office or simply into The Cloud.

The Cloud

This unfortunately, in my opinion, is where things can get a little fuzzy for many people because The Cloud phrase has been misused and misrepresented by those jumping onto the buzzword bandwagon.

The Cloud isn't just somewhere to store your files. That's a file-server which can be located locally or in a data-centre where you drop a file into a box and then fetch it again later from anywhere that's connected.

The Cloud is a resource that includes a computing component with that data storage. Some are more sophisticated than others. But in general its just an off-site computer resource that may be shared, a part or the whole of a real machine.

Again. How does this help us?

Well, here is where things get interesting. Data from production lines, and by implication individual machines, is now available in a central location. Machine data is able to be manipulated in the context of its production line. Multiple production lines may exist at one plant. Multiple plants may exist at various physical locations.

Since we're most interested in Data Analysis and Visualisation, there's a hierarchical way to look at it, from individual production lines to the organisation as a whole:

• Individual production lines.
• By plant / facility basis.
• Company-wide.

For an individual plant we drill down into the interaction between machines and their effect on productivity and efficiency. This is happening in real-time at the plant but incidents that occurred in the past may be replayed and analysed at any time. The production process can be examined for consistency and trends can be picked up.

Comparing the performance of similar machines across production lines and locations is now possible. When it comes to a company-wide analysis similar comparisons can be carried out.

Processes and not Devices

The lines start to blur a little now because we've gone from points-of-interest at individual machines to seeing things at a production line level. And from there, the view of the plant and then across all plants from an even wider viewpoint.

We've moved from the flow of product to the flow of information about the technology that enables production. After all that, we don't want to lose sight of the point of the system which is to get more product into the warehouse.

Improvements in efficiency at the machine level will propagate up to the production line. Lessons learned can be shared across plants. Similar lines from different plants can be compared and analysed.

Processes and procedures now change. Improved processes will facilitate increased efficiency and productivity.

The Road

Improving efficiency with connected devices, a network of things, via The Cloud.

6. MES and the IPAV System

by Dale Whitfield, May 2013

What's the difference?

IPAV is not another Manufacturing Execution System. Its a combined hardware and software solution.

The IPAV System is completely stand-alone. There is no reliance on existing systems or installed hardware. IPAV is installed with its own sensors and IO hardware that gathers data at each machine on a production line. Inputs can be taken from existing PLCs where necessary but this is not required.

The IPAV solution is a combination of hardware that monitors the production process and its connection to custom software. To make this even simpler to install and maintain, the system uses industrial proprietary wireless to communicate between all the installed devices.

Its all about Improving Productivity

Any system like this is all about improving efficiency and productivity. To do that involves various levels of employees in the chain. Not just management. MES is primarily a management tool.

IPAV blurs the distinction between management and machine operator control of production efficiency. Operators can make use of our unique Productive Downtime which lets them know when time is available to stop their machine without impacting on line efficiency. Its that simple.

Operators are at the sharp end. They are the ones who can accurately let management know the Causes for Downtime. We allow Downtime Causes to be customised for each machine, on each line, in each installation. Operators also know how long any particular downtime stoppage is likely to take. We let them enter this time and communicate it to all other operators on the production line. Now the Operators are sharing information and letting each other know that there is additional time available to them for maintenance or cleaning during production.

For management, the state of the production is constantly visible. The the current bottleneck is displayed constantly. The overall state of the Production Line is clearly visible at a glance. This allows higher level decisions to be made that can impact on productivity.

Overlaps with MES

IPAV integrates the areas that cover Business and Operations Processes.

There are overlaps like OEE, Line Efficiency and Machine Efficiency. But in a sense, as important as these numbers might be to some, these are almost a by-product of the system. IPAV calculates and displays these but allows a much deeper level of analysis of the production process.

IPAV allows you to react to problems as they occur. If in doubt, replay it with the built-in record/playback system. Watch what happened in the build-up to where a specific problem occurred and be smarter about dealing with it in the future.

IPAV aims to improve productivity

• It aids in real-time management and tracking of the production process.
• Designed specifically for coupled manufacturing.
• OEE and productivity reporting is only part of what we do.
• Introduce new algorithms and methods for highlighting areas of concern.
• Involves Operators and notifies them of Productive Downtime.

Who Benefits?

MES is primarily a management tool.

IPAV is a tool for both management and operators.

The time taken to react and to communicate solutions to problems is reduced by allowing operators to participate in the decision-making. Operators have immediate and direct control over improvements in efficiency and productivity. There's an incentive for operators because their actions can directly improve productivity.

The benefits are clearly to both management and operators.

Operators handle the detail of productivity by dealing with machine stops and the utilisation of Productive Downtime that results.

Management can make decisions based on the visualisation that is supplied by desktop software.

5. Bringing together TPM, OEE, Lean and 6 Sigma

by Dale Whitfield, March 2013

How these come together with the IPAV System

The IPAV System monitors the production process, analyses it and allows it to be visualised. The intelligent part is the built in knowledge we have of production lines (mostly high-speed), their design and use, and how we manage the data we gather.

But how does that fit in with TPM, OEE, Lean, 6 Sigma and even the Theory of Constraints (TOC).

TPM is a maintenance program philosophy which integrates equipment maintenance into the manufacturing process.

OEE divides the production process into Availability, Performance and Quality metrics.

Lean Manufacturing is all about how to produce more for less. Or how to improve efficiency by means of optimisation.

The 6 Sigma philosophy includes elements which strive to improve the Production Process by removing the causes of its poor performance.

These are different but complementary philosophies with similar goals. We utilise aspects of all of them. Ultimately we want to improve production (i.e. product in the warehouse) with least effort and most transparency.

How do we bring elements of all these together

• Firstly and most importantly, we monitor the production process in the most transparent way possible. We don't rely on existing machine I/O or PLCs. We install our own sensors to monitor product flow.
• Decisions and Indicators are data-driven. We have a lot of data and we use it.
• The operator terminal present at each machine is 'aware' of all other machines on the line. It knows what's happening at its own machine as well as the rest. This allows us to give the operator information based, not only on his/her own machine but on what's happening elsewhere in the production process.
• Then we analyse and allow the Line Manager or a Supervisor to visualise the data. This means that at a glance, any problems with the Production Process are immediately visible.
• Since data about all machines is flowing in real-time for the production process, we can highlight the bottleneck or constraint machine. This highlights the cause of any loss of production.
• Pulling information from all the above allows us to give feedback to allow better Management and Control of the production process.

Feedback has multiple points of influence

• Machine operator: can make use of Productive Downtime for maintenance during production as well as assisting with targetted maintenance by entering the causes of machine downtime.
• Line manager: is able to direct maintenance and make decisions that affect the production process as it happens.

In Conclusion

Targetting and improving efficiency and optimising the Production Process means better performance.

There is no Silver Bullet. There is no one thing that is going to make the Production Process better and more productive. It takes a combination of factors.

By giving operators and managers a holistic view of the process and by creating the feedback loop from machines and operators to the controllers of the process, we complete the cycle that removes the causes of poor performance and thereby improves efficiency and metrics.

4. Productive Downtime meets Total Productive Maintenance

by Dale Whitfield, 19 March 2012

(PD meets TPM)

The IPAV System provides a real-time view of the production line as a whole and continuously calculates Productive Downtime.

Productive Downtime, in the context of coupled manufacturing, is defined as the time available for a production machine operator to stop his machine without affecting the overall performance or efficiency of the production line. See An Overview of the IPAV System for more detail.

Downtime is inevitable. What's not inevitable is that downtime is lost!

The IPAV system allows an operator to indicate to all other machines on the production line how long their machine will be down to resolve an unscheduled stoppage. This we classify as static downtime. In addition, by virtue of accumulators and their capacities, there is dynamic downtime. The combination of these is what we call Productive Downtime.

We prefer to take a positive view on this time, hence the term Productive Downtime.

Total Productive Maintenance (TPM) is a maintenance program philosophy which integrates equipment maintenance into the manufacturing process. One of the goals of TPM is to keep equipment producing with little or no downtime and to keep unscheduled maintenance to a minimum.

Productive Downtime integrates perfectly into this philosophy by indicating to the operator when time is available during manufacturing, to perform preventative, periodic or corrective maintenance.

Rather than adopting the approach of separating out maintenance and production time, Productive Downtime highlights time during production when it is safe to perform maintenance.

The critical factor in this is that Productive Downtime takes into account a machine's impact on Line Efficiency. Productive Downtime indicates how much time is available before a machine will have an impact on Line Efficiency, thereby allowing an operator to choose how best to use that time.

Productive Downtime facilitates the integration of equipment maintenance into the production process.

There are many other indicators that contribute towards TPM, available from the IPAV system.

These include:

• OEE (including the breakdown of Availability, Performance and Quality) on a real-time and shift basis
• MTBF and MTTR are tracked in real-time, per shift and long-term, since the start of equipment monitoring

3. Improving Efficiency by Managing Constraints

by Dale Whitfield, 30 November 2011

Background
The IPAV System of hardware devices, advanced algorithms and software tools facilitate managing the constraints inherent in a high-speed production-line. They constantly build an overall view of the line and analyse various aspects of it. This holistic view and the data it gathers allow improvements in Line Efficiency to be realised.

Managing the Constraint

Managing the Constraint happens in a number of significant ways.
By tracking the location of the Constraint Machine at all times.
By informing the operator of the status of their machine at all times.
By informing each operator of the time available to them before they impact on the Constraint Machine and thereby affect Line Efficiency.

Where is the Efficiency measured?

The slowest machine on the line, whether designed or measured, is the one that relates directly to Line Efficiency. If it stops running, Line Efficiency is impacted and there is less product in the warehouse.

This means that we need to ensure that this machine runs as close to 100% of the available time as possible. In other words Machine Utilisation and Line Utilisation must be maximised. The aim is to keep it supplied with product and ensure that there is always space available for it to discharge completed product into its accumulator/buffer.

Tracking the Constraint?

Tracking the Constraint Machine is performed by evaluating what we call Productive Downtime. The machine with the lowest available Productive Downtime is the bottle-neck and becomes the constraint.

By virtue of Productive Downtime we know where the Constraint Machine is located at any time and are now able to manage it by ensuring that all operators are aware of how they impact on it.

There are two relevant definitions here:

1. Current Bottleneck Machine (CBM) - the machine is the constraint but is producing.
2. Current Constraint Machine (CCM) - the machine is the constraint but is stopped.

Informing the Operator

We are able to forewarn the operator and predict potential impacts on all machines but with special emphasis on the Constraint Machine.

The status of a machine is classified by means of 3 indicator lights:

1. Green - machine has Productive Downtime available and may stop for that amount of time before there is any impact on line efficiency.
2. Yellow - machine is the CBM and will become the CCM if it stops producing.
3. Red - machine is the CCM and is costing line efficiency.

The status indicator light is the first and most visible predictor.

When the light is Red, the machine is stopped and is costing Line Efficiency. All efforts should be made to get it running again, as soon as possible.

If its stopped and the light is Green, there is no impact on Line Efficiency. Available time is being used productively. This is Productive Downtime.

The IPAV Systems Predictor's display shows the available Productive Downtime  in two forms:

• Machine Time - the time available before efficiency of the neighbouring upstream or downstream machine is impacted.
• Cumulative Time - the time before the machine impacts on the efficiency of the Constraint Machine.

These available times are predictions as to when the current machine will impact on the Constraint Machine and therefore impact on Line Efficiency.

What to do with the information

This information allows operators to make decisions on when and for how long to stop their machine. This deceptively simple piece of information can be used in a number of ways which impact on the efficiency of both line and machine.

Impacts on Line Efficiency are straightforward. The aim is to ensure that the slowest machine on the line, at the bottom of the V-Profile, runs as close to 100% utilisation as possible. That translates directly into product in the warehouse. If a machine's status is yellow, it should not stop.

Impacts on Machine Efficiency relate to time available to perform preventative maintenance tasks during production. Tasks that fit into available Productive Downtime can be scheduled will help to eliminate downtime due to breakdowns.

The information also allows management to schedule maintenance tasks that will be performed during periods of Productive Downtime. By carrying out these tasks during production, maintenance days or maintenance periods are shortened. Less downtime means more uptime, Line Utilisation will be greater and with that efficiencies and most importantly, product in the warehouse, will increase.

2. Deriving a Production Line State from Machine States

by Dale Whitfield, 30 November 2011

In our processing we define Machine States and then derive the Line State by inspecting the current Machine States

Generally this applies to machines used in a high-speed coupled-manufacturing environment and we focus primarily on beer and beverage production-lines. However, this technique is not limited to that environment but is limited to coupled manufacturing lines.

Individual Machine States may transition through Sleep,  Startup, Running, Holding, Build-back and  Starving, according to conditions in the immediate vicinity of the machine.

Transitions from one State to another occur as the result of changing conditions either at a Machine or within its upstream or downstream accumulators.

Machine State Transitions

The Line State is derived by inspecting all Machine States at once and as a whole. These states transition through the following.

Line State Transitions

1. An Overview of the IPAV System

by Dale Whitfield, 30 November 2011

What it is and what its not!

• It is a system with the objective of improving and optimising operations and machine efficiency on a coupled manufacturing line by making best use of current assets.
• It is a real-time system which allows the visualisation and monitoring of operation on a production-line in a completely transparent manner.
• It is independent of machine manufacturers and in its basic form doesn't require any physical or electrical interaction with the machine being monitored with the exception of the sensors monitoring product-flow.
• It is stand-alone embedded hardware with high availability and uptime.
• It is not a SCADA or database system that requires kilometres of wiring, server rooms or databases that generate weekly reports in what we call a "post-mortem" fashion.

Theory of Productive Downtime

In general, the concept of Downtime on a production-line is seen as a negative one and is to be avoided. This is always true of the bottle-neck machine, typically the slowest on the line and the one at the bottom of the V-profile, which should strive for 100% run-time. Anything less will affect Line Efficiency. However, for the other machines on the line, there is scope for Productive Downtime.

Introducing the concept of Productive Downtime.

Productive Downtime is inherent in the system operation and utilises a system's constraints, over-capacity and acceleration capability, to predict the location and magnitude of available time at workstations.

It has evolved from production theories like JIT and Kanban which are not applicable in a high speed coupled manufacturing installation with fixed accumulation. Productive Downtime is available due to both this accumulation and stoppages on the production-line.

Coupled manufacturing refers to an installation where machines are connected to each other by conveyors or storage bins known as buffers or accumulators. A machine processes product and feeds it into an accumulator from where the next machine on the line fetches, processes and feeds its accumulator. This continues until the product exits the last machine into the warehouse.

Productive Downtime is applicable to any coupled manufacturing production line whether it consists of high-speed machines or the manual assembly of products.

Productive Downtime Defined

Productive Downtime, in the above context, is defined as the time available for a production machine operator to stop his machine without affecting the overall performance of the production line.

Importance of the V-Profile

High-speed production lines are designed such that machine throughput-rates are lowest in the centre and highest at the start and end. This is the V-profile.

The machine at the bottom of the V needs to be kept supplied with product. In a food or beverage line this is where product and empty containers meet. If it stops, production stops. The longer it runs, the more product enters the warehouse. The IPAV System monitors if the V moves from its designed position and can operate with a dynamic V-profile.

Dynamic Productive Downtime

Upstream of the V, a machine that keeps its accumulator filled is able to create Productive Downtime by virtue of the fact that its downstream machine has product to draw on and therefore stay running. Downstream of the V, machines draw product away from the V-machine to ensure it has accumulator space for its product. These create dynamic Productive Downtime.

Becoming the Constraint

Any machine that doesn't fill or drain the V-machine will become the Current Bottleneck Machine (CBM) or Current Constraint Machine (CCM). We define the CBM as the machine with the lowest Productive Downtime on the production line. This is a good indicator that the machine is not performing well. A machine transitions from CBM to CCM when it stops producing. When there's a CCM there is a direct cost to production line efficiency.

The Productive Downtime real-time advantage

Where the IPAV System has an advantage over traditional methods is that it allows the state of the production-line to be monitored, analysed and displayed in real-time at each machine. Existing technology adopts a post-mortem approach. Stoppages are analysed after the fact.

This allows the operator to make decisions based on the current state of the production-line. The operator is aware that if his machine is the CBM, stopping it will make it the CCM and impact on line efficiency. Maintenance or cleaning decisions are made based on available Productive Downtime. These all combine to improve overall line efficiency.

Static Productive Downtime

When a machine stops due to operator intervention, emergency stop or breakdown, the system prompts the operator for the expected downtime. This is static (breakdown) Productive Downtime. Productive Downtime at any machine is the sum of its dynamic and static downtime.

When static downtime is entered, the time is propagated to all other machines on the production line. Since the breakdown machine is expected to be stopped for the given time, all machines benefit. The system monitors MTBF and MTTR and can use this to broadcast estimated downtime without operator intervention.

Productive Downtime is an Advanced Warning

The IPAV System predicts Productive Downtime in four different line states: Sleep, Start-Up, Running and Run-Out. See the posting on Line States for a detailed discussion on how these states are derived.

Many production lines do not have line-of-sight communication between machines. When the line has emptied of product for example, i.e. Run-Out and gone to a Sleep state, each operator has an indicator showing when product is expected to arrive when it restarts. Depending on the line configuration, this could be an hour or more from the first machine starting to run, to product arriving at the last. Operators can use this time more effectively than if they had inaccurate or no knowledge of the state of the production line.

IPAV Systems Equipment and Installation

IPAV Systems equipment at each point-of-interest consists of a Heartbeat device which gathers all sensor signals and a Predictor which displays Productive Downtime to the operator.

The line manager has the option of a Globe device which gathers data from the Predictors and displays a summary of the current state of the line. The Globe has an Ethernet connection and serves up web-pages to any desktop PC browser.

Using the Globe's Ethernet connection, the GlobeLive cross-platform desktop application, supplied free of charge, provides detailed graphical analysis, monitoring and visualisation of the production-line in real-time. A truly powerful tool.

Installation of the IPAV System is simple and non-invasive. Each point requires power and sensors to count product entering and leaving the machine. Communication between Predictors uses robust proprietary wireless adapted for industrial environments.

Additional Features

The IPAV System incorporates many features beyond the scope of this article.

• Extensive statistics and performance indicators for machines and operators.
• User and pre-defined messages may be broadcast to other Predictors and Globe and by SMS to pre-programmed phone-numbers. 
• Data for each machine is stored continuously, e.g. 121 days at 5 minute intervals and retrieved via USB or Globe.
• Line Configurations may be changed by uploading a parameter file generated on a PC desktop using a spreadsheet or the GlobeLive application, allowing product changes with varying pack sizes to be set up quickly.
• Simulation Mode provides a sophisticated mechanism to model a line and experiment with changes before deployment into the field.
• GlobeLive application supports detailed analysis of individual machine parameters.
• GlobeLive application supports recording and playback of the activity at each machine. For example, record and then playback a shift to analyse performance or any other parameters stored. Manual selection of recording is possible and the last 3 days of 24hr data are recorded by default.