Don’t underestimate the risks!

cpp: Mr. Feucht, all the signs are pointing to a paradigm shift in the automation industry with regard to wireless. Following the fieldbus revolution, should manufacturers of measuring instruments now be getting ready for a wireless revolution?

Feucht: I hope not! I must admit, I find it difficult to share the euphoria this topic sometimes arouses. Wireless solutions obviously have a few key advantages because they do away with the wiring. That’s a particular benefit when it comes to temporary installations or controlling moving parts of machin-ery and equipment. The lower costs and reduced effort are other strong arguments in favour of wireless if sensor data needs to be acquired in applications that are not time-critical. In spite of this, I believe that extensive wireless networks to control processes or plants are dangerous and often unrealistic – not to mention the impracticality of coordinating workflows in real time.

cpp: What exactly are your objections?

Feucht: Wireless solutions can never provide the same level of safety and security as wired systems. Wireless transmission is by definition extremely susceptible to electro- magnetic interference – a simple jammer is all it takes to bring it to a complete standstill by suppressing the carrier or overriding the receiver. This has the effect of directly limiting availability and ultimately impairs safety. A functionally safe system that is forever changing to a safe state because communication breaks down is no good to anyone. If that’s the case, experience shows that sooner or later the facility operator will look for a way to bypass the safety mechanisms and render the functional safety at least partially ineffective. And bearing in mind that the available bandwidth is restricted, I rather doubt that this kind of solution is sustainable. The wireless systems can no longer be guaranteed to be interoperable. After all, the free license bands – what we refer to as ISM – are in the process of being regulated right now. I could cite several current examples where existing regulations have been revised again subsequently – to the consternation of those operators who have already upgraded their plant with wireless technology. That’s not what I call sustainable! There’s also the threat from lobbying interests at odds with those of automation specialists. In the end, compared to the big wide office world, automation is no more than a niche.

cpp: In your opinion, what impediments exist for standards with something to please everyone?

Feucht: When fieldbuses first appeared in the market back in the eighties, there were umpteen different technical approaches, and it took a lot of time and nervous energy to specify a technology pool in the form of IEC 61158. The situation is no different today with wireless communications. And each of the many, often very heterogeneous wireless technologies – WLAN, Bluetooth or Wireless-Hart– claim purports to be the perfect choice for industrial applications. Once again, the path to standardising potentially suitable technology candidates is a rough one. What’s more, the circle of interest groups is now much bigger than it was with the fieldbus bodies – just think of the recently revised ETSI 300 328, for instance, or the draft of the ISA SP100.11a standard for IEC 62734, which has just been rejected in a ballot.

cpp: Isn’t it normal for standardisation processes to take a while?

Feucht: The real problem is not the time it takes to reach a consensus but, as I already mentioned, the underlying interests of the manufacturers, which at best only partially coincide with the central interests of automation companies. Wireless technology was primarily conceived as a standard communication technology for consumer and office applications. The potential market is huge, which means we’re dealing with a giant international community of interests. Automation industry representatives tend to play only a minor role. As a result, their influence on the allocation of frequency bands or the formulation of telecommunication standards – like ETSI EN 300 328 – is minimal. Users of radio-based real time solutions might as well pack up and go home at 100 mW EIRP.

cpp: What does this standard actually involve?

Feucht: ETSI, the European Telecommunications Standards Institute, is planning to introduce various technical interoperability measures from 2015 in response to the increased use of the 2.4 GHz band. The aim is to prevent the growing number of devices operating at this frequency from interfering with one another. One objective of the new stand-ard is that before a device starts to transmit it should first check whether any other devices are transmitting on the same channel. If so, it must either use a different channel or wait. Unfortunately, these regulations won’t just create extra challenges for designers and make the technology significantly more expensive; they’ll also slow down data transmissions. The measures undoubtedly make sense for home and office applications, where it simply isn’t possible to organise wireless networks in a way that ensures full interoperability. The delays they provoke are almost imperceptible and can generally be tolerated by users. It’s an altogether different situation with industrial applications, however, where time is a critical factor. Deterministic data communications with millisecond resolution are not compatible with delays on this scale. At 100 mW EIRP this affects WLAN, Bluetooth, Z-Wave, WirelessHart and many more besides. In other words, these wireless stand-ards will in future only be suitable for controlling and monitoring relatively leisurely processes.

cpp: You also mentioned security. Where do you see room for improvement here in wireless communications?

Feucht: One of the fundamental weaknesses of wireless is that anyone can access the transmission medium owing to its very nature. Although total encryption protects integrity and confidentiality, that’s only one half of the story because the transmission itself can be interfered with or suppressed by very simple means. The fact that absolute availability cannot be guaranteed is the single biggest shortcoming of wireless. Of course, wired data communications are never one hundred per cent safe from interference either. That’s why we have detailed EMC directives stating precisely which radiation and irradiation limits apply to devices in different applications and why we go to such tremendous effort to shield cables, terminals and devices. And quite apart from the consequences of unintended interference, willful tampering also can’t be ruled out. Whereas it takes a great deal of time and effort to sabotage wired communications or infiltrate malware, however, wireless transmissions can be seriously disrupted – if not brought to a complete standstill – by a jammer, possibly from a considerable distance. As an active amateur radio operator, I speak from experience. Yet, even if we assume that there is no evil in the world, almost every technology enthusiast today has a story to tell about overlapping frequencies interfering with consumer electronics equipment or causing it to be operated incorrectly. As I pointed out before, the ETSI wants to overcome the problem of limited frequencies by revising the standard but is omitting to take account of industrial practice. That could turn out to be an expensive mistake, as we saw when the LTE licences were auctioned off to wireless operators to the detriment of camera and stage equipment.

cpp: Why is most of the criticism from industry ignored?

Feucht: Many manufacturers are evidently unaware of the implications of these problems, and in many cases they don’t even realise there is one until they’re directly affected. That’s the only explanation I can think of for the enormous enthusiasm for what they call Industry 4.0 and the almost unbounded trust in the security mechanisms of data communication technology.

cpp: Why are you so much against the Industry 4.0 concept?

Feucht: I could talk for hours on that subject but basically I think it’s overrated; it’s a case of old wine in new bottles. We already have various manufacturing execution systems that make production processes incredibly flexible and allow even small batches to be coordinated easily. If Industry 4.0 means each workpiece contributes its own machining parameters themselves, as it were, the price will inevitably be even more complex data communications and increased use of wireless and Internet technologies. The broad standardisation of the communication technology and the protocols this concept entails will be enough to open the gateways even wider for any attempts to paralyse production. What especially worries me is that the age of software attacks on production facilities is only just beginning and there’s definitely no reason for complacency. Responsible minded decision-makers will have to face up to the truth that technology has its limitations. It doesn’t matter in the slightest how long a particular technology works without any major problems. The litmus test for the process and manufacturing industries is what happens if ever a problem does occur. For developers of failsafe systems like Knick, a healthy distrust of technology is just as vital as engineering expertise. Nevertheless, I’m convinced that the natural instinct for what high tech can achieve and what it can’t is on the retreat among many manufacturers and users – and with it their awareness of the risks. The popular definition of sustainability today is that the material from which a defective device was made must be suitable for recycling. The consciousness that a long service life, investment security and durable products are equally important aspects of sustainability has evidently been lost. The process industry traditionally insists on these attributes and it’s not uncommon for system solutions to remain in use for twenty or thirty years.

cpp: That sounds as if what you’d really like is a complete renunciation of wireless solutions.

Feucht: No, of course not. Back in 2006, Knick launched a ZigBee-based system for transmitting pH sensor data. There are some applications where wireless unquestionably makes sense. I simply want to warn against the erroneous conclusion that wireless systems could maybe revolutionise the automation industry. Wireless is an efficient and affordable alternative whenever failures are unlikely to have any serious repercussions. If the aim is to monitor parameters in non-time-critical processes and deterministic data communications are not a must, I don’t see any problem. If the worst comes to the worst, the person on duty in the control room can always get on his bike and cycle over to the measuring point that’s causing the hiccup.

cpp: But what would be your recommendation for critical applications where wiring is not a viable option?

Feucht: Careful coordination within the plant in combination with radio relay. The normal rubber aerials radiate into the surrounding area like a sprinkler, so that hardly any of the original transmitting power actually reaches the receiver. At the same time, the highly sensitive receiver input eagerly picks up any form of interference. Twisted pair or coaxial cable remains the only practical choice.

cpp-net.com/0213437