Look, no wires!

Although wireless devices – especially the mobile phone – are now central to Western civilisation, they have yet to become as firmly embedded in industry. Wireless telemetry, sensing and communication are all employed in manufacturing, but at far lower levels than might be expected. “Much of the reason for this is conservatism – and a fear that there are still security and reliability issues with wireless devices,” says Khadambari Shanbagaraman, research analyst at market consultancy Frost & Sullivan. In a report on the benefits of wireless technology to factory automation, she points out that the automotive industry is leading the charge to adopt the technologies – driven largely by the need to gather real-time data. Some of the other benefits of wireless technology include: remote monitoring ability; ease of installation; and reduction in cabling cost (leading to savings of around 30%). But the many perceived hurdles, including a lack of awareness and the relatively high investment costs – are still putting people off. Another issue preventing adoption is confusion over the many varying standards – and a fear that a new system will not fit with existing architectures. “Manufacturers need to see the benefits of these systems if they are to invest in them,” she says. “In many cases, it’s not worth replacing an existing system. Wireless systems will be implemented when new facilities are built.” Wireless software Software developers have also taken note of this. National Instruments, for example, has enhanced LabView 8.6 with a wireless sensor interface – allowing it to take advantage of multi-core processing and gather more test data at higher rates, without having to install masses of wiring. The wireless technology come in the form of software and hardware, with capabilities similar to that available on most laptops, giving such systems a range which marketing manager Ian Bell describes as “hundreds of metres”. In LabView 8.6, the capability to programme in a wireless interface has been added, without the need for code changes in existing data acquisition applications. Bell says: “We also have support for other wireless networks, such as Zigbee, but not in our hardware.” Companies such as Saphir in France, which specialises in signal acquisition and processing, are already reaping the benefits. It has created a noise and vibration monitoring system for building sites that can monitor structural health and capture noise pollution information above legal limits. “With the flexibility of LabView, we were able to re-use our existing code to quickly expand our measurements using Wi-Fi data acquisition devices,” says its president Jean-Michel Chalons. Closing in for the kill But for electronics design house Plextek, the lack of take-up is more simple: wireless sensing still needs to come up with the ‘killer app’ that will create a market winner. The company recently led a study into wireless sensor networks (WSNs) for communications industry watchdog Ofcom – and concluded that the technology is still restricted to a core number of applications. Steve Methley, senior consultant at Plextek who led the research, said: “The lack of a killer application may be due to limiting factors such as the current cost of wireless nodes, and a lack of understanding and experience by end users.” “One likely sign of a movement towards a killer application would be the involvement of a major systems integrator,” said Methley. While ‘unlicensed’ portions of the spectrum can easily support WSNs, a dramatic increase in use could change this – with bands becoming crowded with data. “This may lead to the appearance that WSNs are unreliable – an especially important issue, as the perception of unreliability is one of the key barriers identified for WSN adoption,” said Methley. Plextek believes that WSNs will become more widespread over the next three to five years as systems continue to adopt licence-free bands. For now, the applications taking advantage of wireless networks are traditional sensing applications – such as industrial and environmental monitoring, building automation and remote sensing. “Usage will continue to evolve in these areas as experience grows and the benefits of reduced installation costs are increasingly understood,” said Methley. He said growth of WSNs also relied on improvements in battery and ‘energy scavenging’ technologies. In the latter category, UK company Perpetuum – which featured in Eureka’s August 2006? issue – has recently completed a trial to gather data from wireless sensors in a gas plant. The company’s ‘microgenerator’ – which harvests energy from machine vibrations – together with a supercapacitor from Cap-XX, was used to power the sensor system. Cap-XX vice president of applications engineering, Pierre Mars, said: “Wireless system manufacturers can now easily design battery-free systems using this self-generating power source.” Wireless warning A very real problem of wireless networks – and one that is hindering its greater adoption – is reliability: too many sensors reporting too much data can too easily fall over. There is also a danger that individual sensors can be damaged – which is a problem within a factory, but a nightmare for true remote sensing, such as in the natural environment. Winsoc, a project involving European and Indian researchers, aims to put that right. It will mimic living organisms in order to design ‘self organising’ wireless sensors for environmental monitoring – that are robust enough to overcome the failure of individual sensors. “The starting point was to provide mathematical models of biological systems and translate them into algorithms to determine how the sensor nodes should interact with each other,” says Sergio Barbarossa of the University of Rome, who coordinates the project. “Living systems are intrinsically robust against cells dying or being damaged.” The principle has been tested in two ways: in a computer simulation of a forest fire; and as a prototype network in India, as a landslide detection system. A network of 12 geological sensors – connected to 15 wireless nodes – has been installed in a rainforest in Kerala, which is vulnerable to landslides during the monsoon season. The sensors monitor the humidity and porosity of the terrain, as well as acceleration forces. The sensors are linked to a satellite, which gathers the data and conveys it to a control centre. Barbarossa believes that self-organising wireless networks will be very important in environmental monitoring. “A system capable of autonomous decisions is particularly important in remote areas – where it is difficult ot recharge batteries or replace defective nodes,” he says. Stand-by for savings A wireless sensor network will sit at the heart of a project that aims to save energy in office buildings. According to TWI, which leads the Lesstricity project, around 20,000GWh is consumed by electrical devices that are left on stand-by – or 6% of the UK’s domestic electricity consumption. TWI and its project partners, including GSPK Design, will develop a new platform technology that can detect electrical equipment that has been left switched on. The wireless network will cause minimum disruption, while being easy to install. “It will operate as an independent system or can easily be integrated into a building management system,” says programme manager Nee Joo Teh of TWI. The research is supported by the Technology Strategy Board. Pointers Automotive is leading the adoption of wireless technologies in factory automation – mainly because of the need for more real time data * National Instruments has developed a wireless sensor interface with a range of hundreds of metres to cut down on wiring times and costs * Wireless sensing still needs a ‘killer application’ if it is to turn a corner, according to Flextronics