The military uses for wireless technology are persuasive. For example, pilots can fly above a war zone and drop thousands of small wireless sensors, the size of a small pebble and costing a dollar apiece, over the terrain. As soon as they settle the devices start communicating with each other, weaving themselves into a dense digital mesh. They pick up vibration and sound, so they can identify advancing troops. The sensors can also detect the presence of nuclear, chemical or biological agents. The information they pick up is relayed to a satellite. For power, they “scavenge” energy from the environment, using solar energy or temperature changes.
The civilian uses are equally impressive. Forest rangers can drop the sensors from aeroplanes to detect fires, showing their exact location and how fast and in what direction they are spreading. Smaller versions the size of grains of rice can be used by airlines in the innards of aircraft to check for the presence of large insects or rodents that might interfere with the wiring. Still smaller versions the size of specks of salt can be added to paint, turning entire surfaces into wireless sensors that can detect motion or act as smoke alarms or security systems.
The trouble is that as yet such sensors do not exist. But a lot of practical work is already going into making them a reality. A version of the military scenario above, for instance, was rehearsed in an experiment in March 2001 at an American Marine Corps base in California. Around a dozen nodes the size of a matchbook were released by a miniature unmanned aircraft. They were able to measure the speed and direction of vehicles from ground vibrations. This proved that the technology, although still at an early stage, was viable.
The military experiment was supported by America’s Defence Advanced Research Projects Agency (DARPA) and arose from a programme called “smart dust” at the University of California at Berkeley in the 1990s. It created some of the technical foundations needed for sensor networking, such as a pared-down computer operating system, database and protocol for sensors to send traffic (but “sleep” as much as possible to prolong battery life). The technologies are open standards, so may be used freely by other firms, just as the underlying protocol of the internet is open.
With DARPA’s early support a gaggle of companies have emerged, such as Dust Networks, Arch Rock and Moteiv (from research at Berkeley) and Ember (based on work at MIT). Other companies, such as Crossbow Technology, Millennial Net, Sensicast, Tranzeo and MicroStrain, are applying the innovations to their existing technologies. Wireless-sensor technology is now moving out of military testing grounds and into the commercial world. It is used for things like monitoring and controlling industrial machinery, automatic temperature regulation in buildings and keeping tabs on the environment.
Whereas M2M communications generally involve wireless devices attached to equipment such as cars or vending machines that link up to the cellular system, sensor networks use small chips that are often embedded into the device and use a local-area network that may never connect to a larger one such as the cellular system or the internet. For the moment the sensors are not yet widely deployed: the technology is still maturing and customers need convincing that it is worth having. But the idea is gaining momentum. Once the volume goes up, prices will come down and follow-on innovation will speed its adoption.
A good place to get a glimpse of new wireless technologies in action is BP’s Cherry Point Refinery in Blaine, Washington. Built in 1971 on almost four square miles (10 sq km), it has a daily throughput of 225,000 barrels of crude oil. The site also produces 8 percent of the world’s calcined coke, which finds its way into one out of every six aluminium cans. Modernising the plant to keep it efficient is costly; BP says it has spent nearly $500m on this over the past ten years. Fields of tanks need to be monitored for operational, safety and environmental purposes. But snaking wires across such a huge area is expensive.
New wireless technologies are critical, explains Tim Shooter, who works on future technology at BP. They allow more operations to be monitored and controlled and save money at the same time. The average refinery has around 3,000 “instrumentation points” where data on things like temperature, flow, humidity and vibration are collected; managers would be even happier with 10,000 points if only they were less pricey. The cost of the basic monitoring devices ranges from $1,000 to $10,000 apiece. Although adding wireless functions to these sensors almost doubles that cost, it reduces the price of installation by 50-90 percent — and installation makes up most of the total cost. By upgrading some processes to wireless systems, Mr Shooter believes each refinery will be able to save at least $1m a year.
Getting Better All the Time
Until a few years ago wireless technology was not up to the job. The “big leap forward”, says Mr Shooter, is that the new technologies are far more reliable in hostile industrial conditions and the communications protocols are more intelligent.
One notable innovation is “ad-hoc mesh networking” in which each node on the network — eg, a sensor on a water pump — is both a transmitter and a receiver and can join the network whenever required. Earlier wireless technologies assumed that sensors would send data to a specific receiver, in a hub-and-spoke fashion. This had many drawbacks. For a start, it made the system inflexible. If you added a new node, the whole system had to be reconfigured and the network became harder to manage. And if a central receiver failed, the whole system collapsed.
The newer technology remedies these faults. Each node can relay traffic to other devices, creating an interlocking web. It needs less power because the data travel only a small distance to another node. It is self-organising and self-healing. If one node goes down, the system finds an alternative path for the traffic. And the more devices are attached, the more efficient and resilient the network becomes.
Factory controls like those at BP are obviously useful, but they do not add up to a volume business. A good example of a large-scale application is building management. Thanks to wireless communications, lighting, heating and air-conditioning can be controlled centrally to keep energy bills down. So when a guest checks out of a hotel the receptionist can adjust the air-conditioning to stop it needlessly chilling the furniture. This could also be done through a wired system, but wireless technology offers lower installation costs and greater flexibility.
Some firms are installing such systems in older buildings as well as integrating them into new ones. Riga Development, a wireless-technology firm in Toronto, has worked with hotels in Canada and the United States to replace ageing analogue thermostats with digital ones that are around 35 percent more energy-efficient. It wirelessly links the new temperature-control panels with heating and air-conditioning units, at a cost of around $350 per room. Each room can also be controlled from the front desk. And thanks to the wireless mesh network, the panel in each room also acts as a relay for the data traffic from other rooms back to a central control point.
At a medium-sized office park in Las Vegas, wireless temperature controls were installed in a few buildings containing around 200 offices, says the media-shy maintenance manager (who did not want his or the company’s name to be used). Temperatures in the Nevada desert tend to extremes and landlords are responsible for energy bills, so managing a building’s climate makes a difference to the bottom line. The new wireless thermostats allow rooms to be controlled centrally on a PC or over the web. The adjustments that tenants themselves are able to make can be controlled too, so that heating or air-conditioning is not used to excess. The system was cheap to put in, mainly because it required very little installation, the manager explains. Tenants are happier and the savings on the energy bills have been considerable, he says; “conservatively 25 percent”.
These uses of wireless are just the beginning. Sensors are not only being added to devices that already have electronics on them, but being put on to things that were formerly bare of any technology at all. For example, they are being fitted to buildings, bridges and roads to monitor their structural integrity. The sensors can identify stress and early cracks that need attention. Sensors are also being used to monitor the environment. Scientists now use them to measure the climate in areas that would have been impractically small when sensors were more costly — say under individual plants rather than in a thicket.
Wireless sensors are also cropping up on farms, to measure temperature, moisture and light on tracts of land where wired sensors cannot easily go. Among the first big users are vintners, because their crop is particularly valuable and even small variations in climate can ruin it. Ranch Systems, for instance, supplies equipment and software to a dozen vineyards in Northern California. A fleet of sensors allows growers to monitor wind, water and soil and air temperature. This helps them set the watering schedule to suit the different needs of each part of the vineyard and manage frost, disease and pests, explains Jacob Christfort, the founder. “It is a little like da Vinci and the helicopter,” he says, referring to the artist’s famous sketches that presaged later inventions. “These things were conceptually possible all along, but some mundane advances are required before it all comes together and somebody actually does it.”
The technology has become so accessible that it is sparking a cottage industry of small entrepreneurial firms. Moteiv is putting sensors on firemen’s uniforms to relay information about the fire and let their colleagues know exactly where they are. The system can even provide the firemen with information such as floor plans, projected onto their masks. Other applications now on sale include wireless home-security systems and wireless beacons for sailors to tell the crew when someone has fallen overboard.
Yet the very diversity of its uses highlights one of the barriers to the development of the technology: they all have to be put together in a bespoke fashion. Wireless technology is so new that it has yet to be simplified and standardised, as most technologies are over time, notes Monica Paolini of Senza Fili Consulting.
Another complication is that nobody really knows how much stress a collection of wireless sensors will put on a network, other than that it will probably be different from what happens on the internet. Much internet traffic is asymmetric, with computers at the edge of the network receiving hundreds or thousands of times more traffic than they send. A single mouse-click to request a file brings a massive YouTube video in return. With sensor networks this traffic asymmetry is inverted: they send far more data than they receive. Although each individual consignment of data is tiny, they add up. And some sensors send out a steady heartbeat, if only to say “I’m still here!”, which sets off a communications session throughout the network.
What worries engineers most is how to deal with all the data produced by the sensors. “The good news is that you can get all these data; the bad news is that you have to do something with them,” says Kris Pister, the co-founder of Dust Networks. Efforts are under way to increase the processing power of the sensors so that they can analyse the information themselves rather than just collecting it and passing it on.
But this wealth of information creates opportunities as well. Teruyasu Murakami of Nomura Research Institute believes that having things continuously connected to a network will open up new markets and new ways of living. And Bob Karschnia of Emerson Process Management, which designs and builds factory automation systems such as the one at BP’s Cherry Point Refinery, digs through the mountains of data to find new ways for businesses to operate. At times, he philosophises about what the technology means. The interconnected machines are akin to the brain’s neural pathways, he suggests. “If we are computing and connecting like the brain, we should be able to emulate memory,” he says. “How do you create memories’ in the processes of a factory?”
As machines talk to other machines, they may uncover facts and relationships that are not apparent to people. That may enable factories to “learn” and find ways to become more efficient. What happens on the factory floor will make its way, in a different form, to office buildings and homes. The next step is for wireless technology to enter human beings themselves.