What is the future for safety networking?
Industrial networking protocols have had an excellent year. The tail-end of the recovery from the 2009 global recession resulted in the installed base of networking nodes (including both fieldbus and Ethernet) growing by over 14% from 2010 to 2011, an extra 31 million new nodes.
Increasingly over the past few years, the major protocol associations have been developing protocols to tackle specific problems, safety being prominent among them. It has now reached the point where almost every available protocol seems to have a safety sub-protocol or an alternative; some specialist safety companies are also releasing their own. Are these protocols being adopted as fast as they are being developed?
For new factories, there are several advantages to installing a safety network instead of the more traditional hardwired systems:
- Cost – Fewer cables and fewer components like safety relays are required; with some automation networks, the same network infrastructure can be used for all automation and safety functions.
- Easy Maintenance – Rather than having to check and replace multiple safety-monitoring relays, many problems can be flagged electronically and logic systems can be put in place to allow continued running if the problem is not critical to safe operation.
- Monitoring – Most safety networks can automatically keep track of the number of times a safety function has been triggered and give active readouts of the condition of individual components and the likelihood of failure. This allows pre-emptive rather than reactive maintenance.
- Flexibility – With component auto-detection software on some protocols, components can be switched out with minimal programming. Networking also allows for relatively simple expansion of the facility simply by adding further branches to the backbone.
It seems surprising, therefore, that in a recent IMS Research user survey, fewer than half of respondents had any intention of adopting a safety network within the next 3 years. This survey was carried out only in EMEA and the Americas, but I believe that this percentage would have been lower if Asia Pacific had been included. So why the lack of interest?
A large part of the problem, especially in EMEA, is that very few new factories are being built. Upgrading legacy systems can be expensive and this cost is hard to justify if current safety systems comply with the legislation and are working correctly. The staff trained to maintain and use these systems do not necessarily have experience with networking; and training or hiring simply adds to the cost. One of my peers recently described manufacturers as having an “inherent conservatism," a little harsh perhaps, but I think the fact remains that any new technology trying to enter the industrial market does have a barrier of trust to overcome before it is widely adopted. With an application as critical as safety, potential users are going to want to see proof that these networks are effective, reliable and work in real-world situations before they start replacing hardwired systems they know and trust.
In much of Asia, it is a different story. While a large proportion of the global greenfield sites are in this region, industrial safety is not treated with as high a priority in many countries. While this is slowly changing with the adoption of safety legislation by China and India (amongst others) these guidelines are not always policed effectively; without enforcement, legislation cannot be effective.
These factors go some way to explaining why a recent study from IMS Research found that hardwired systems had a much larger installed base of connections than any safety protocol and that this was expected to remain the case until at least 2015. Figure 1 shows forecasts of installed nodes. Of the safety networking protocols the three most widely adopted are CIP Safety, PROFISAFE and AS-Interface Safety At Work; they are also some of the longest established and most widely promoted by their respective protocol associations.
Overall, safety networking is growing, although not as fast as total networking on a global level. I believe that once the barriers of trust, the perceived cost-benefit ratio, limited experience, and lightly enforced safety legislation are overcome, safety networks will ultimately become firmly rooted in factory automation. The safety networking market is still young and depends heavily on legislation; when legislation is enacted and enforced is difficult to predict. I believe that safety networks will not be generally accepted before 2017, and perhaps much later; but this opinion is open to debate.