Throughout most of the telecommunication industry, manufacturing automation has flourished.
Here’s just one example of successful automation: fusion splicing. The operator preps the fiber ends, puts them into the splicer, and presses a button. The equipment measures and aligns the ends, checks the cleaves and performs the splice tests for quality - all hands-free. This automated fusion process offers enhanced cost savings and improved quality.
But let’s talk frankly about the typical connectivity process. We recognize the incremental improvements over the past 40 years that include increased volume, from polishing a handful of connectors at a time to seventy-two, and automation, from hand pressure technology to mass polishing machines. That said, except for very large-scale manufacturing, the current process looks very similar. We still manually assemble connectors and manually process them. To reduce manufacturing costs, our industry has migrated the manual production of fiber optic cable assemblies primarily to low-labor-cost regions of the world.
So what’s the next step? How do medium and large independent cable assembly houses continue to reduce expenses while increasing connector quality? Can our industry develop “Ready to automate” tools and equipment that can take advantage of existing generic automation processes and software and yet still provide for the rapid change and flexibility needed in the connector assembly space?
We, as an industry, must come together to agree to address automation in fiber optic connectivity. In fact, I believe automation is an essential next step for the fiber optic cable assembly industry. However, we face multiple challenges. The following paragraphs discuss several of these challenges, followed by my proposed solution – a way for our industry to overcome the obstacles and take a giant step forward.
Challenge: Stable technology
One of the concerns with automation in the optical fiber industry is that you want a stable technology. Again, fusion splicing serves as a good example: when companies began using fusion splicing (as opposed to other methods of permanent termination) that process became very stable. Whether factory or field splicing, they are using the same equipment and stability. The point is that while one process was developed (fusion splicing) the process equipment continued to upgrade and add more features but the core process did not change. We had one process, initially very operator sensitive that due to the advances in generic robotics and manufacturing automation allowed automation. An example from a different industry is the assembly of automobiles. For decades, auto manufacturers have robot-welded unibodies together. This stable technology successfully automated a major non-stable step in the manufacturing process.
We cannot stop progress – and we don’t want to – which means no industry is immune to changing technology. (Compare Ford’s Model T to the self-driving Tesla) When we look at the history of connectors and connectivity, “progress” has meant the continual development and refinement of different styles and types of connectors in order to lower costs and improve quality and ease of assembly. Fiber optic cable assembly initially used adapted copper coax connectors (SMA), then ceramic ferrules, non-optical disconnects, then multi-fiber connectors with many different styles and types (LC, SC, FC, ST, MTR, MTP, etc.). Then we split out again moving from only, factory-installed connectors and adding field-installable connectors, many of these depart from traditional manufacturing techniques to finished connectors with a splice-on area in the rear and others that are actually installed in the field and hand-polished.
Together, all these variations represent an ever-improving technology – technology that is sometimes disruptive. So when we talk about automating the connectivity process, we have to first look at automating the in-factory assembled mass-produced connectors. (As an aside, I’m fairly certain some large manufacturers have automated the processing of specific types of high-volume connectors, with highly proprietary procedures.)
Challenge: Flexibility in automation
Note: all connectors of a specific type must meet an interminability standard know as a focus document. This document defines the mechanical standards that allow different manufacturers products to work with each other. A copper analog of this is the RJ 45 Ethernet copper connector. So therefore each manufacturer can and does have different internal designs and innovations. Add to this the fact that we have literally hundreds of cable and breakout ends to mate to these connectors and you will see the need for a very flexible manufacturing process in this landscape.
In addition, flexibility in automation is critical, because we don’t want to stifle future technological advancements. Every year, we see new innovations in connectors. For example, there have been a number of variants of so-called ferrule-less connectors. Other types of connectors have metallic, ceramic, or molded plastic ferrules. Plus, some connectors actually provide a bare-fiber-to-bare-fiber interface through a gel. A large number of these variants have come and gone or found specific niches.
There are two key reasons our industry has not been automating the production of connectors:
- We’ve been able to reduce the cost of making the basic connector bodies. Much of this relates to automated injection molding processes and ceramic ferrule standardization.
- We’ve been able to reduce the cost of labor by moving the large connectorization operations to low-labor-cost regions. That migration has actually stymied the market for medium and large-scale, flexible automation that can deal with the rapid pace of technology changes in connecterization.
It’s interesting to note that our industry has accepted and required sophisticated testing and inspection processes and has brought them to a very high level of Island automation. By that I mean that the equipment provides process automation testing within its own environment but is not neutrally interfaceable with other similar equipment. We have not brought the actual connector assembly operation to that level of automation. Cable Assembly houses continue to struggle with a bottleneck in process at that testing stage and eagerly anticipate improvements.
Challenge: Enhance cost-savings and quality
In order to move the automation of connectivity to a large number of manufacturers, there has to be a reason. The reason is this: As we go forward, every connector must offer lower loss and higher reliability. The connectorization process must produce fiber optic cable assemblies with increasing levels of quality while decreasing cost.
As noted above, a major challenge is dealing with so many different connector designs and materials. More important, how will connectors be designed in the near future? We don’t want to stifle innovation while lowering cost and creating higher reliability.
Yet we need to face the fact that today’s assembly process is basically a very manual process in the vast majority of companies – a manual process with severe limitations. A colleague is fond of saying, “A human is a two-sigma being. No matter what you do, your reliability and quality is generally at two-sigma performance.” Yet we need to move to a three-sigma or even six-sigma performance. Why? Because a higher level of performance is absolutely critical in certain applications. Examples include life/safety issues such as flight critical connectors in aircraft, connectors in laser surgery applications, and connectors used in 911 emergency or military communication networks.
Definition of Sigma:
Sigma is a unit of measurement about statistical significance; the standard deviation expressed with the lowercase Greek letter sigma (σ). It is the amount of variability in a given set of data regardless of how close together or spread apart the data points are.
Normal distribution are the results of an experiment that when plotted on a graph, creates a shape highest in the middle and tapering on each side. This is often referred to as a bell curve.
Deviation is how far a given data point is from the average.
Standard deviation is the square root of the average of all the squared deviations.
The 68–95–99.7 rule is a shorthand used to remember the percentage of values that lie within a band around the mean in a normal distribution with a width of two, four and six standard deviations.
One standard deviation, or one sigma, plotted above or below the average value on that normal distribution curve, would define a region that includes 68 percent of all the data points. Two sigmas above or below would include about 95 percent of the data, and three sigmas would include 99.7 percent.
A five-sigma result is considered a gold standard, corresponding to about a one-in-a-million chance that the findings are a result of random variations. A six sigma is considered one chance in a half-billion that the result is random. The well-known business strategy called “Six Sigma” is derived from this term and based on mandating rigorous quality-control procedures to reduce waste.
In order to go to that next level of factory automation for fiber optic connectivity, we need to think in terms of islands of flexible automation, much the same way manufacturers in the electronics industry have embraced flexible automation. Think about the ever-growing complexity of the printed circuit board assembly business. That industry moved from basic printed circuit boards to multi-layer circuit boards to surface-mount technology. Now they are developing chip-to-chip technology. Each step of the way, the equipment that was designed for the previous generation of technology became surplus, and the old process migrated to lower-labor-cost areas. Eventually, each technological process was outmoded – sometimes in fewer than 5 years.
So the questions here are:
- How can we simultaneously embrace automation and not stifle new technologies?
- How can we deal with changing requirements and specifications?
- How can we implement and pay for automation in a short period of time?
Some automated technologies can serve as a guide. For example, the 5-axis and 6-axis robotic arms were designed and built to very flexible (literally), so they can be easily updated over the years. Several industries adopted standardized software models for automation interfaces so different manufacturers components worked together seamlessly. This is an insurance policy, so to speak, so the technology doesn’t become quickly outdated.
In order to deal with the ever-increasing and lower cost requirements of the fiber optic cable assembly industry, we must work with the assembly equipment manufacturers to come together and agree on automation process interfaces both software and hardware related.
In fact, I believe end customers will drive the demand for increased quality and flexibility in assembly. The escalating demand for three-sigma to six-sigma performance will force cable assembly houses away from the “test-the-quality” way of working to a “build-in-the-quality” approach.
When it comes to automation in connectivity, our industry is at a tipping point. It’s time for action.
In the cable assembly process, our industry has successfully automated advanced testing and inspection processes. However even here we do not see easily aligned automation protocols that allow for different manufacturers systems to seamlessly integrate. We now need to automate fiber optic connectivity. As a representative of the finest manufacturing companies supplying cable assembly houses, I know that many of these suppliers also see this need. However, no one manufacturer has been able to take the lead. This stands to reason, since each supplier is an expert in its own technology, created for a highly specific step in the process.
What’s the solution? I believe the quality suppliers in the fiber optic industry must come together specifically to address this urgent issue. Once we fuse together a consortium of manufacturers – an automation working group They will need to look at how other industries have tackled this challenge and note available software and hardware conventions that will need to be decided and adopted for the assembly equipment manufacturers to move forward. These AD Hoc choices may lead to further standardization and this can only help the industry. This is a tall order, but it’s clear that the time is NOW.
The question is: Can we agree on such automation standards to tackle this industry issue? Can we interest automation software and hardware companies to assist us in this undertaking? Also, can we be flexible enough to deal with the different cable configurations, connector configurations, and ongoing technological upgrades?
Automation in fiber optic connectivity is a pressing topic for the cable assembly industry. Let’s start the conversation, and let’s move the issue forward. This topic must be addressed now – it’s the essential next step for the fiber optic cable assembly industry.
Additional resources from the FOC team include:
- Category Resource
- View the Glossary, Acronyms, Military Specifications for Connectors
- Q&A Resource: email technical questions to AskFOC@focenter.com
Have questions about this article?
Contact FOC with questions at: (800) 473-4237 / 508-992-6464 or email: FiberOpticCenter@focenter.com and we will respond ASAP.
