Go big or go home? The pros and cons of small batch assembly in EMS

small-batch-assembly-chemi-1080x675

It’s become something of a ‘small batch’ legend.

In his classic introduction to ‘Lean Thinking’, James Womack narrates the tale of how he challenged his children to an envelope-stuffing competition.

They decided to set up an assembly line: one of them folded the letters, another labelled and placed a stamp on the envelope and a third stuffed the newsletter in each and sealed it.

James opted for a small batch approach: he folded, stuffed and stamped one envelope at a time.

As in all the best fairy tales the small batch hero conquered the over-sized, giant assembly line.

And the lesson of the story is:

Henry Ford only got it right some of the time.

Small batch assembly

We’re so used to bowing to the gods of economies of scale and continuous production that we can easily become blind to the moments when these are very much false idols.

Think for a moment about why James won his challenge (and, no, it wasn’t because he was the adult).

Stuffing one envelope at a time got the job done faster because of all that hidden time it took to sort, stack and move around the large piles of work-in-progress envelopes on the assembly line.

While it’s certainly true that each child became more efficient at their allotted task, individual performance was, in the end, less important than the overall performance of the system.

The benefits of small batch assembly for OEMs

Let’s take a look at small batch assembly and consider where it may offer an attractive alternative to OEMs.

  • It allows you to be more agile

    In these days of increasingly customised, and occasionally personalised, devices, small batch assembly allows you to quickly switch between the production of different models and variants.

  • It allows you to be more responsive

    Identifying and fixing quality problems can often be done much sooner and more cost-effectively, as you don’t have to wait until the end of the assembly line to discover problems.

  • It involves much less up-front and overhead investment

    To achieve the economies of scale of long assembly runs an investment in raw materials and components in bulk is needed. As well as the increased inventory costs for input materials there is also the resultant abundance of products at the end of the run that offer overhead costs for storing.

  • It can reduce manufacturing risk and get you to market quicker

    Small batch assembly allows you to get a finished product into your customers’ hands faster. Particularly for prototypes this is critical – enabling you to validate, gain feedback and implement any specification changes that might be required. It also allows you to trial new products without committing to holding large stocks until sales performance is confirmed.

The drawbacks of small batch assembly

Like all fairy tale heroes, small batch assembly is sometimes flawed.

The majority of time (depending on product complexity) goes into setting-up a surface-mount manufacturing process. Running a batch double the size, effectively amortises the set-up cost twice as far. In this case, there is no substitute for volume: bigger is better.

Apart from losing out on cost for longer runs, the main failing is that unless a small batch assembly is meticulously planned and optimised to reduce downtime and cycle time, it can very quickly become inefficient and expensive.

Thankfully, with practice and expertise this, at least, can be avoided. But, the very model of small batch assembly requires for a certain amount of inefficiencies as equipment is stopped, re-configured and tested before the next batch can be produced.

Continuous run or small batch?

Having sung the praises of the small batch it’s worthwhile taking a reality check. For longer runs in greater volume there is no doubt that continuous production is cheaper. But there is a place for the small batch.

We always advise customers to let us help them understand the issues, priorities and ideal deliverables for each product (or range of products) and build a risk-averse plan based on this.

Undoubtedly there are huge benefits in creating scalable volume, but it’s all about planning and careful management. A combination of small batch assembly segueing into continuous production can often offer the best way to progress to volume and optimise your manufacturing costs.

And, hopefully, we’ll all hit those profit figures happily ever after.

The IoT upgrade: Exploring what IoT means for manufacturers and the supply chain

In a recent Gartner report, the research and management advisory firm predicted that there will be 25 billion Internet of Things (IoT) enabled devices on the market by 2020.

The opportunities the IoT offers OEMs are vast.

IoT enabled products can:

  • Optimise and improve manufacturing processes
  • Add automation, performance and personalisation to consumer products
  • Enable more smart military capabilities
  • Develop more intuitive healthcare services

And that, of course, is just the tip of the iceberg.

Yet there are some notable IoT challenges for OEMs and risks that still stand in the way of success.

Understanding IoT for electronic products

IoT is frequently misunderstood as a ‘thing’ or a certain set of technologies that can be applied to a product. In fact, IoT’s implications for supply chain decisions and product design go much deeper than this suggests. The IoT is actually best understood as a massive leap in the evolution of technical communication.

Imagine a large group of people attending a party. They all speak different languages and the handful of translators the host has kindly supplied are really struggling to keep up with the demand, let alone the multiple potential combinations of conversations requiring translation from one language to another.

The result is slow, stiff, awkward, largely unproductive and uninspiring interactions, which make for a very dull party.

And now imagine how things would be different if the host had developed a new language that all the invited guests could speak in common, allowing them to interact with and understand each other with ease.

Suddenly everyone is able to converse, be amused, entertained and learn from each other. The effect is explosive and dynamic, offering an even better fizz, bubble and sparkle to the party than even the most potent punch bowl could.

The opportunities for interactions are now huge, if slightly chaotic. This is a party that could potentially head in many new directions and take several surprising turns. A party which needs the right levels of organisation and resources to work properly, but in which the possibilities are exciting and potentially endless.

This is the IoT party – and you’re invited.

The IoT upgrade

Electronic products and devices that have been dancing to the old beat now need an ‘IoT upgrade’ if they are to join the fun.

But this will not happen by simply adding something to existing products. It will require a new way of thinking, a new way of designing and to take advantage of the latest manufacturing capabilities to succeed.

The IoT upgrade starts from the ground up, and it’s important that it does not occur in isolation.
The growth in IoT and embedded connectivity needs to, and is, taking place alongside a number of factors. These include:

  • A continuous drive for ever greater miniaturisation of electronic products
  • A market demand for constant innovation and ever-enhanced product functionality
  • An increased need for electronic product designers and manufacturers to be fast in reaching the market through faster and more efficient procedures
  • And competition levels that demand extremely tight unit costs can be achieved

The IoT upgrade itself has the potential to upset long-established market dynamics, and it takes place in a global supply chain climate that is extremely fluid and fast moving.

OEMs who succeed in this climate will have developed the ability to maintain a flexible, responsive and agile production operation. And critical to this are the partnerships they forge with key EMS service providers.

IoT challenges

There are many challenges for OEMs undergoing the IoT upgrade.

There are potential issues with the power requirements of edge devices that will typically be required to operate from small batteries. The choice of low-power requirement connectivity solutions can greatly aid in maintaining product size requirements.
There is also a very fine balance to be struck between allowing for rapid innovation but also ensuring that everything is safe, secure and rigorously tested.

The supply chain has already created significant issues for some OEMs as lead times have been compromised when sudden growth in popularity around certain devices or technology has led to a shortage of parts or components. As more OEMs undergo the IoT upgrade we can only expect this to become a greater challenge.

There is also the issue of cybersecurity. Big brands have already suffered hacks, notably Ford Chrysler’s and Tesla’s infotainment systems, and V-Tech whose IoT toys were targeted leading to more than six million children’s personal information being compromised.

When considering medical, defence or aerospace OEMs and products, security has to be of utmost importance. This is best achieved using a ‘security-by-design’ approach where devices are built from the ground up to be secure, rather than having security features added after they have been delivered and deployed. It is also a factor that can be affected by the choice of protocol used (such as open, close, digital, analogue and so on).

The IOT upgrade and early EMS design engagement

The challenges faced by OEMs undergoing the IoT upgrade can be minimised and managed by early engagement with manufacturing and supply chain partners.

EMS design teams can advise on suitability and availability of parts, components and techniques that each product’s manufacturing process can be best served by.

This early engagement ensures that expensive discoveries or delays caused at a later development stage are avoided.

To discuss the design, manufacture and sourcing of your IoT upgrade, call the Chemigraphic team on 01293 543 517.

Is winter coming for Fieldbus?

It’s been nearly 20 years since the Fieldbus war ended. A ‘Memorandum of Understanding’, the closest industrial automation has ever come to an Entente Cordiale, was signed on the 16th July, 1999, by some of the loudest voices in the debate including Siemens, Rockwell Automation and Fisher-Rosemount. This effectively put an end to years of fighting over whose version of Fieldbus should be adopted in the draft IEC 61158 – the European standard that aimed to set parameters for industrial communication networks.

The large and comprehensive IEC standard was released the following year, accommodating all competing Fieldbus systems and eight different protocol sets called “Types” were agreed. For the companies involved in the memorandum, it was worth the battle to protect these investments, as the industrial communication market is now expected to reach USD 141.0 billion by 2023, according to a MarketsandMarkets global forecast.

industrial-communication-market

However, the many competing technologies for Fieldbus meant the original hope for one, single unified communications mechanism was never properly realised. In part, this was simply not practical, since Fieldbus technology needs to be implemented differently in certain applications, e.g. automotive Fieldbus is functionally different from process plant control, but also the development paths differed. Despite each technology sharing the generic name of Fieldbus, the various technologies are not readily interchangeable. The differences between them are so profound that they cannot be easily connected to each other.

While the Fieldbus market has been divided, Industrial Ethernet has been steadily marching onwards, and many believe that one day it will sweep aside many of the Fieldbus systems. In fact, recent studies of the industrial network market by HMS Industrial Network reveal that Industrial Ethernet has overtaken traditional Fieldbuses in terms of new installed nodes in factory automation. With a growth rate of 22%, Industrial Ethernet now makes up for 52% of the global market compared to 46% in 2018, the study claims.

Is this the end of series battle that Fieldbus cannot win? Before we attempt to answer this question, let’s go back to the basics.

Explaining Fieldbus

Fieldbus is a way to connect and control in real-time a network of field devices – sensors, actuators, motors, switches, drives, etc, within a manufacturing plant. Self-evidently, it comes from two terms, Field and Bus, where the Field represents the plant levels and the Bus is the connections (circuits) between units which transfer the data. Prior to Fieldbus, point-to-point parallel wiring was used to connect devices, but increasing amounts of automation necessitated higher levels of wiring complexity and cost, which also led to more human error. Fieldbus gradually replaced these point-to-point links between the devices and their controllers (such as PLCs, CNCs etc) with a digital single link on which all the information is transmitted in serial.

Fieldbus

Source

A set of rules is defined in order to accomplish data transfer between the units along the bus – called the protocol. The protocol is responsible for two important rules on the bus, the mechanism that any unit can acquire or seize the bus (from the network terminology this means the way of Medium Access), and the synchronisation between those multi-units on the bus.

Devices are wired together using a single ‘trunk’ cable (typically a twisted wire pair carrying a digital signal and DC power) and communicate via the relevant Fieldbus interface and software. As the Fieldbus communicates only via a cable, it has been possible to decrease the wiring considerably in comparison to parallel wiring.

What are the alternatives?

Ethernet is widely used in home and industry. The internet protocol is commonly carried over Ethernet and so it is considered one of the key technologies that make up the Internet. As such, it’s everywhere. As the name suggests, Industrial Ethernet is the use of Ethernet in an industrial environment with protocols that provide determinism and real-time control. Ethernet’s steady march over the decade has found its way into various protocols. Of them, EtherNet/IP is now the most widely installed network at 15%, followed by PROFINET and PROFIBUS, both at 12%, according to the HMS Industrial Networks report.

So is winter really coming for Fieldbus?

The majority of industrial components has historically been connected through serial Fieldbus protocols, such as PROFIBUS, Control Area Network (CAN), Modbus and CC-Link. PROFIBUS is the world’s most successful Fieldbus technology and is widely deployed in industrial automation systems.

The Fieldbus army is therefore still formidable and the systems are still indispensable within the industry.

However, Industrial Ethernet has been growing faster than traditional Fieldbuses for a number of years and has now overtaken them, offering higher speed, increased connection distance and the ability to connect more nodes. With a growth rate of 22%, Industrial Ethernet now makes up for 52% of the global market compared to 46% in 2017, according to the report.

One significant difference between Fieldbus networks and Ethernet networks is speed. Standard Ethernet protocols achieves very low latency and enable a flexible network topology. As such, many industrial communication protocols are moving to Ethernet-based solutions. The real-time accuracy for Industrial Ethernet protocols, such as PROFINET and Ethernet/IP, is more accurate – about 1ms. This ensures specific manufacturing data is sent and received on-time – exactly when the data is needed to perform specific operations. Fieldbus systems were designed for the typically noisy plant environment, and as a result speed needed to be slower and of course a suitable physical media found, especially for those protocols in which data and power are in the same single cable.

Complexity and costs has been one of the major causes of decline. As a Fieldbus represents a complete system, qualified personnel are required for its operation. The individual Fieldbus components are also considerably more expensive.

The way forward

The original aim of the introduction of the Industrial Ethernet was unification of the communication infrastructure from the guide level to the sensor/actuator bus. Unfortunately, exactly the opposite occurred: at present, the number of Ethernet solutions developed until now is even higher than the number of Fieldbuses. This makes a comparison of the different systems nearly impossible for the user.

The simple answer to the not so simple question is that technology in this (ahem) field is constantly evolving and, what may work for one sector or application, may not for another. The upside to this vague conclusion is that there is a multitude of options available, each suited to different sectors, requirements and applications. Manufacturers and designers must align themselves with experts not only in the technology but in the manufacturing processes used throughout production to select the best option to suit their individual needs.

It might not be the Mother of Dragons, but fieldbus may have legitimate claim on the title of Mother of Communications Interfaces?!

Turning Ideas Into Action: Valuable NPI Lesson for Tech Start-ups

‘Hold on a minute – I’ll just use my Apple Newton to take a message.’

Not something you hear every day. In fact it was never something you were likely to hear because the Apple Newton, an early Personal Digital Assistant (PDA) device which was launched in 1993, didn’t sell. A combination of factors, some of which included its high price and early software problems that hampered the handwriting-recognition feature, prevented it from ever gaining acceptance or popularity.

So it’s true – even Apple can fail at New Product Introduction (NPI). Google has suffered many relative failures too – Google Glass was a flop and Google Hangouts was adopted by some but not by the many, as with the majority of its other services. They could’ve been, they should’ve been – but ultimately, they weren’t.

There are several important things for tech entrepreneurs and visionaries to take into account as they stand on the edge of this exciting and challenging phase of the production process.

High Stakes NPI

When you’re a trillion dollar company you can afford a few mishaps and the occasional bad decision, but what happens to those tech start-ups who have already gone from idea to prototype but fail to introduce their products correctly? An NPI failure at this level usually means the company collapses and everybody down the line – entrepreneurs, angels, crowd-funders and, sadly, employees too – all lose their investment and maybe even their livelihoods.

It’s important to remember that NPI is a very high-stakes part of the process and it’s absolutely critical to get this right. It’s invaluable to have an experienced manufacturing partner who will typically introduce several hundred prototypes into production every year, to advise you through the process, as there’s many steps that go into successfully launching a new piece of hardware. This can range from schematics, prototyping and sourcing, all the way to mapping quality processes and returns processes. As we’ll see, the NPI process is arguably one of the most crucial elements of electronics manufacturing and possibly the most challenging and complex one. That’s why we are committed to providing our customers with a smooth transition from design phase to manufacturing and finally out to market.

Return to prototyping

‘We’re done with this. We need to be first to market!’ Of course you do, but are you sure the product will be accepted? Is that what Apple said about the Newton? Perhaps it’s worth getting a few more people to try them out…just to be certain?

Nobody likes to second-guess themselves too much, but the more people you test your product out on, the more you’ll see them interacting with it in ways that you never thought possible. Watching a user try to destroy your product can actually be very satisfying – especially if they succeed (sounds counterintuitive but it will give you the chance to rectify a problem). You may learn more in a few minutes of observing someone interact with your product than in the years you spent in building and developing it.

Don’t Rush – speed to market isn’t everything

 It’s important to remember that consumers and businesses are now more hungry than ever for new electronics that are lighter, faster, smarter and more connected, and you’re always competing with other manufacturers to introduce new technology. However, whilst staying competitive is key, a race to failure is not one you want to win. Lead times are constantly being squeezed in the race to produce market-first products. This poses a big challenge where NPI engineers have to develop tighter than ever controls and processes in order to deliver products which not only satisfy quality, function and robustness, but also deliver speed to market.

Take DfM (Design for Manufacture) seriously

Your prototype might be functional and working, and it might even be manufacturable, but has it been optimised for volume manufacturing, especially if there’s a sudden, high take-up? If your demand curve takes on a ‘shark fin profile’ will your prototypes work within a high-volume manufacturing environment? That’s the crucial test.

DfM is the practice of designing products with the manufacturing process in mind, choosing the best processes, materials and components. DfM addresses this by asking designers to consider not what could be created, but what should be created. For example, have you minimised the use of new ‘active parts’ in favour of standardised, widely available components, as the design of a new part is usually only the best option from a purely inventive design point of view. 3D printing/additive manufacturing techniques are invaluable in the prototyping stages, but are the tolerances too wide for volume production? Are components being selected that come in machine-friendly formats for fully automated processes?

When a customer involves us early they can optimise the product properly using DfM principles. It’s worth remembering that decisions made during the design phase determine 70% of the product’s final cost.

Don’t overlook the supply chain

Product design engineers sometimes struggle to see beyond the immediate prototype or small-batch production stages. We see them focussing on functions and components, which is obviously very important but without fully understanding the challenges of obtaining some of these parts, especially at the moment with some components like MLCC (Multi Layer Ceramic Capacitors) in short supply. Even the most perfectly designed electronics assembly, presenting zero fabrication, regulatory or inspection issues, can create critical delays and costly substitutions if components are not sustainably available. It’s essential to conduct an availability analysis to determine how easily parts can be obtained as well as their cost. At this point, any potential issues identified in the supply chain can be addressed. Having access to BOM management tools that will identify parts in short supply and suggest alternatives is where an EMS partner can add significant value.

Accept that your designs may need to change

We are sometimes presented with a shiny new prototype and asked for the costs of volumising production. It might be the best design in the world but it remains a difficult question to accurately answer because the prototype may still need to change.

New product designs typically go through two building verification phases: engineering verification and design verification. The initial engineering verification is how we test the inner workings, especially of printed circuit boards (PCBs). Once the initial PCBs are validated, additional testing systems are employed as we extend the testing into areas including durability, safety, and reliability. The next build, design verification, will be much closer to how the finished product will look. It’s important that start-ups remain open to product evolutions throughout the NPI process.

The Chemigraphic difference

While NPI is exciting and rewarding we’ve seen that products can fail at this stage if great care isn’t taken to get things right before the green light is pushed. The importance of having a skilled manufacturing partner at your side cannot be overstated. By engaging with customers early on in the design process, we can optimise product designs to reduce cost, make processes more robust and aid the transition through the prototype and production phases.

Our NPI early engagement and engineering team collaborates regularly with our customers on all aspects of design, component sourcing, assembly and test to create an optimal package. If you want to see a good example of our NPI process in action, see here