Wireless Sensor Networks - We Are Getting There
2010年5月11天 Dr Peter Harrop
One hundred years after Faraday, came that other European giant of electrical theory and experiment, Nikola Tesla, who also parted his hair down the middle. Faraday brought us the dc motor and much more besides: Tesla brought us the ac motor and much more besides. However, Tesla did most of his work in the USA and became an American citizen. In 1905 he foretold that we shall be communicating with hand held phones all over the world. Another hundred years onward and the progress towards that dream has been truly awesome.
Yet it is only a first step to small devices communicating without human involvement and without those radio masts and their expensive and vulnerable cabling. Remember that cellphones like land lines went down during the Haiti earthquake. It is not just East Asian hackers take down our systems: they remain extremely vulnerable and that includes the internet with its hard-wired infrastructure. Just power outages from a volcano or an earthquake can take them down.
Mimicking the way a message is passed through a crowd of people, so-called ad hoc networking of radio signals is on its way. In its most general form it is like a mesh, where anything can communicate with anything provide it is near enough and ready enough. Like people hearing and talking, mesh networking devices can both receive and transmit without infrastructure. For example, after three decades of development, we now see tens of millions of utility meters in large buildings communicating via these mesh networks using the inherently meshed ZigBee protocol and derivatives. Start up Meraki Networks connects 400,000 San Francisco residents to the internet via their Free the Net ad hoc networking project and others aim to replace the tangle of wires in the home with mesh networking.
Wireless Sensor Networks WSN are machine to machine M2M mesh networks operating like the internet in that they are self organising and self healing. If one "node" is out of action, then the message eventually gets round by another means. Drop them from a helicopter: they do the rest because they are "self-calibrating". Little wonder that battlefield communications is a primary proving ground of WSN and they will be used to study thermal vents at intense pressure on the seabed because they tolerate node failure.
We should like these new networks to handle huge amounts of data and be completely maintenance free so we can put them on a billion trees to monitor forest fires, seal them in the concrete of buildings or the metal of engines and even drop them on disaster areas such as oil spills to monitor every animal and slick and help to deal with them. The SNCF national rail system in France already finds that they survive well in rolling stock.
However, as with people passing on a message in a crowded room, there are severe problems of maintaining such nodes if they are to be there for long.Batteries must survive and remain charged or something better must be used. There are severe problems of interference and garbled messages in both cases as well so the mathematicians, standards writers and systems analysts are working on that. Wireless bandwidth is a constrained resource, knowing node location is troublesome, coping with moving nodes is difficult and there are other challenges.
With WSN we therefore compromise - hugely. In a world of real time communication, WSN often gives very delayed messaging. The message only passes when another node comes within range and power is available, perhaps because the sun has hit the solar cell on the node and charged the supercapacitor or the heat of the engine has provided thermoelectric energy.
There is therefore a trend towards putting two or even three different kinds of harvester in each node. Alternatively, a lithium thionyl chloride primary battery can last 20 years in the node but its use may have to be minimal- not too many messages passing and not for too long. Tadiran batteries leads here.
So far, the packets of data are extremely small with all WSN, but Twitter has shown us that small packets of data can achieve a great deal. So have EnOcean "No Wires, No Batteries, No Hassle" radio-emitting light switches and other controls in over 100,000 buildings. Some are two way but not yet meshed. EnOcean relies on energy harvesting by electrodynamics with a miniature version of your bicycle dynamo capturing movement - thank you Faraday. They also use photovoltaics and thermoelectrics. Add piezoelectrics used by others and all are leading us to the batteryless WSN, though we are not there yet: not in volume use anyway.
In that crowded room, sending a message is easier if people have strong voices and good hearing. The electrical equivalent is chips that use less power in the WSN node. Huge advances have been made here. In the picture we show the power needs of various electrical and electronic devices and how WSN have been "gas guzzlers". Some discharge their batteries in only a few weeks. If they have to rely on energy harvesting into rechargeable batteries we have to recognise that no rechargeable battery is guaranteed for even ten years and their life is often far less.
The good news is that companies such as Dust Networks now have WSN routing nodes taking a mere 90 to 300 microwatts and leaf nodes at the edge of the network functioning on a mere 30 microwatts. Such figures match those of traditional active RFID which is much more primitive in its capability. Suddenly tiny broad band vibration harvesters, transparent film photovoltaics and many other forms of harvesting become sufficient and the 20 year node looks possible.
A forthcoming event uniquely pulls all this progress together and airs the exciting applications that result. It is "Energy Harvesting and Storage Europe/ Wireless Sensor Networks & RTLS Europe" in Munich Germany with a two day conference 26-27 May, optional masterclasses before and after and visits to local centres of excellence and an exhibition. Presentations from GE, NYK, EADS, Continental, Northrop Grumman, the US military SPAWAR research center, the aeroengine research of Rolls Royce , ST Ericsson, Texas Instruments, Philips, Dust Networks, Schneider Electric, the ZigBee Alliance, EnOcean. Tadiran Batteries and many other world class developers and users show the way.
In due course, the new printed electronics, a revolution as significant as the microchip forty years ago, gives the prospect of memristor electronics mimicking the synapses of the human brain. This could mean even lower power consumption. We may have harvesting, electronics and sensors all printed on top of each other on plastic film or paper at very low cost and, where necessary, biodegradable and/ or transparent devices. For example, the nodes monitoring that oil slick and its trapped animals will decay into harmless material when their task is done.
It is certain that there will be other major breakthroughs that we cannot currently envisage, including perhaps by a third genius after the third century following the other two. Maybe that man or woman will also part their hair down the middle like Faraday and Tesla. Stranger things have happened.