Improving turns in machine building

Management at turnkey automation systems providers are always seeking ways to improve production efficiencies on the manufacturing floor. The challenge is to keep up with sales growth and deliver systems on time, while maintaining the same amount of resources.

Machine turns or simply ‘turns’ is a very powerful measure of productivity for turnkey automation systems providers. Simply put, turns measure how many machines can be built in a year in the same space with the same amount of resources. For example, if a 10,000 m² facility can produce five machines simultaneously, then in other words, production of one machine takes up 2000 m². If each machine takes four weeks of production time, then the same 2000 m² could produce about 13 machines in a year (assuming 365 days of operations for simplicity of calculation). Therefore, the machine turns of this space is 13. The total production capacity of the manufacturing floor would be 65 machines a year. Now, if the machine builder decides to increase production to keep up with growing sales, there are two options: add facilities in multiples of 2000 m², maintaining the same turns on the floor, or increase turns with existing resources. If this machine builder can save four days per machine it would improve turns from 13 to 15, making the facility 15% more productive, compared to investing more capital and resources for an additional 2000 m² to achieve 20% more productivity. The relationship between turns and efficiency can be seen in Figure 1.

The concept of turns is not new and neither is the concept of process improvement. There is a plethora of knowledge on improving manufacturing efficiencies — from adopting newer technologies and improving internal processes through lean manufacturing initiatives like six-sigma to building sustainable supplier relations for just-in-time production. But what is often overlooked or ignored as an area of efficiency improvement is the controls architecture.

Figure 1: Relation between machine turns and efficiency. For a clearer image click here.

Controls architectures, being the integral part of the system or machine, are usually not thought to be an area for process improvement. Of course, as with any process improvement initiative, cross-functional teams are necessary. In this case, engineering and production play an important role.

Cabinet-mount philosophy versus machine-mount philosophy

The traditional (and most prevalent) approach to managing low-voltage sensors and actuators on a machine is based on a centralised control strategy — also known as cabinet-mount philosophy. In this approach, while sensors, valves and other electromechanical devices are out on the machine, the wires controlling their actions are routed back to the controls cabinet, a long distance away, where the PLC or controller for the system resides. Once the wires are routed through the maze of wire-ways or cable trays, the wire ends are stripped, crimped, labelled and then screwed into the terminal blocks located on the panel of the control cabinet.

Based on the complexity of the machine, one machine could have 50–200 wires or more, of all different colours and characteristics. This requires a significant amount of detailed engineering and planning, not only for the I/O, but also for the routing and wiring diagrams for inside the cabinet. Furthermore, exhaustive debugging follows the electrician’s work of wiring the cabinet. With this approach, I/O mapping and programming become sequential activities.

This centralised I/O strategy is a labour-intensive and therefore time-consuming activity, costing valuable machine build time. Alternatively, a distributed or machine-mount control strategy eliminates several of these steps to simplify the overall machine design and build process. At the foundation of the distributed architecture is the network or the fieldbus system that allows for exchange of I/O as messages amongst networked nodes. Today, almost every machine has adopted some level of distributed controls architecture. A good example of this partial adoption is in robotics. Robot controllers and end-effectors usually have fieldbus or network nodes. The information about hundreds of I/O is communicated amongst the robot, the end-effector and the machine controller over a single fieldbus cable. This tremendously reduces the complexity of the system.

Today’s distributed modular architecture eliminates wiring and extremely simplifies the control cabinet. Additionally, enhanced versions of network I/O blocks today, offered by some vendors, have onboard diagnostics for connectivity, short-circuit protection, and overcurrent protection. This diagnostic functionality saves valuable time during the commissioning of the system. Machine-mount IP67 versions offered by some vendors have an added advantage for deployment in harsh industrial environments.

Benefits of IO-Link

The distributed architecture becomes even more attractive when it is combined with IO-Link technology. IO-Link is a vendor-neutral and fieldbus-neutral communications protocol for point-to-point communication. This protocol is specified by the IO-Link consortium and published in the IEC 61131-9 standard. There are three major benefits offered by IO-Link technology with a distributed controls strategy:

1. Modular machine design with increased I/O and reduced cost: In the most primitive form of the full distributed architecture, each network node can host up to 16–32 I/O points. The IO-Link enabled network blocks could go anywhere from 136 up to 480 I/O per network node. For some machines this can be a lot more I/O than required and offers the benefit of built-in flexibility for the future and the ability to handle any last-minute change requirements with much less effort compared to the traditional cabinet mount case.
2. Labour savings by simplifying complex connections: Valve connections, such as SMC or Festo valve banks, typically require 16–25 conductors to handle expandable 16–24 electromechanical pneumatic valves. With the traditional cabinet mount strategy, installing a single pneumatic valve bank could take 3–4 hours of labour. Conversely, with IO-Link valve connectors, the install time takes only minutes.
3. Reduce engineering and design time with smart sensor integration: Smart measurement sensors such as pressure, temperature, distance or inclination measurement, vision or colour sensors, and even RFID read/write heads are increasingly being used in today’s automation. No matter which vendor supplies these devices, as long as they are IO-Link capable, they can be easily integrated in the distributed controls architecture schema.

Figure 2: Traditional versus IO-Link. For a clearer image click here.

Implementing a machine-mount control strategy

A machine-mount strategy for the controls architecture is not only a huge timesaver during the machine build phase, but also reduces time in tear-down, transport, rebuild and commissioning. Machine builders that adopt the distributed machine-mount strategy often find opportunities for further improvements in the machine. For example, standardising a machine design with a distributed architecture means that programming and configuration tools, along with engineering designs, are easy to replicate on the next machine, thereby reducing design engineering time on the recurring machines to about 50%.

Identify the need

As pointed out earlier, there may already be some level of distributed architecture in your system. To determine whether your system needs a full distributed strategy, look for these signs:

• Inspect the machine build and rebuild schedule: If your machine build schedule includes more than 25% of the time for building control panel and wiring, there is a good chance you can significantly reduce that time to about 1/3 by switching to a distributed controls strategy. It is also ideal to look at the total labour hours for electrical technicians.
• Review the control cabinet: In some cases, especially in building standard machines, the control cabinet build is outsourced, but the final wiring might be completed in-house. This might appear like a good idea because the activities can be done in parallel. However, if you can save approximately 80% on cabinet space and 50–70% on labour time, does that activity still make sense?
• Discuss with controls engineers: Controls engineers are the most affected by changing over to a distributed strategy. Two important questions to ask your controls engineers are: “How often during the machine programming do you refer to the electrical wiring diagrams and perform testing and debugging for the wiring?” and “If you had Add-On-Instructions (in case of Rockwell controllers) or function blocks (for most other brands of PLCs) for all the network blocks and connected devices, would this reduce any installation time?”

Answers to these questions will provide a good understanding of where the time is spent with the current controls strategy.

Get your team on board

As with any improvement initiative, a team’s buy-in on the decision goes a long way in the implementation stage. Design engineering, controls engineering and production teams are primary stakeholders in this decision-making process. The highest amount of resistance may come from design engineering as changing over the control strategy involves significant upfront work. When implementing a distributed modular architecture strategy for the first time, it may take a little longer for the engineering side of things, as it is a paradigm shift, but the savings on recurring machines could offset the cost at a rapid pace. Allocating more time for the first system can work in everyone’s favour in the long run. The significant benefit for the design engineering team is that future customer-driven changes can now be handled with minimal effort.

Find the right supplier partner

Finding the right supplier is an essential step to your implementation success. It is often unclear whether the step of finding the right supplier should be done before or after getting your team on board. The right supplier may help demonstrate the value of the distributed modular architecture to your team or, if the team is already on board, they may ask the right questions to help identify the right supplier. Three important things to look for in a supplier when adopting the distributed modular strategy are:

• Open architecture portfolio: Support for an open architecture offers the flexibility to use the same components on any major network with fewer changes to the bill of materials or the configuration, regardless of network/fieldbus or device vendor choices.
• Strength of technical support or expertise: The best way to understand technical support expertise and knowledge is to request the supplier’s local technical experts to demonstrate the configuring components on your choice of network, possibly on your existing machine. Or even asking the experts to demonstrate the technology by walking through the configurations with your controls engineers could validate the strength of technical expertise.
• Breadth and depth of product line: Understanding the breadth of product support for different networks, industries or applications might provide better insight into the supplier’s ability to respond to market needs.

Tracking time savings

Distributed modular architecture offers several benefits to end users and machine builders in terms of modularity, flexibility and scalability of the system. However, this article strictly focuses attention on the significant time savings potential that eventually results in a more efficient manufacturing floor, without adding more resources.

There are three major areas for time savings potential:

• Labour time in wiring and building the cabinet.
• Controls engineering time in programming and commissioning.
• Design engineering time on recurring machines with minimal changeovers.

The time savings on design changes may be realised over a period of time. The controls engineering time changes are challenging to track and validate — primarily because the requirements are sometimes fluid and their extra time can be utilised to improve or enhance machine functionality. The labour time savings, however, are easy to track and quickly evident from the project timelines.

Figure 3: Comparing labour and component cost.

Example calculations

Let’s review the hypothetical example presented at the beginning of the article. In this scenario with the traditional cabinet-mount strategy for the controls architecture, each machine took 28 days to build. Let’s assume that 30% of the build time or 8.4 days are allocated to the electrical wiring of the cabinet. Changing over to a distributed modular architecture could save between 50–70% of the electrical wiring time. So, 50% would be 4.2 days of savings or 70% would be 6 days of savings per machine. This implies improved turns between 15.3 and 16.5. In other words, the same floor with the same resources can now build 76 to 82 machines instead of the original estimate of 65 machines a year.

*Shishir Rege is Marketing Manager for Networking and Safety for Balluff Inc. and works out of the Balluff headquarters in Florence, KY. He has over 15 years of experience in robotics and automation in diverse industries including automotive, packaging, aerospace and medical.