Enter the maze

Is the grass always greener?

A sheep grazes on a hillside

Sheep in the Lake District might soon be in for a shock. If they don’t keep their eyes open when they’re eating, they may end up munching up some tiny sensors scattered throughout the hills where they live. A national team of researchers, led by Professor Joe Sventek at Glasgow University, are exploring how to ‘seed’ whole areas of our natural landscape with hundreds of tiny battery-powered computers that sense aspects of the world around them. Each computer has a radio transmitter, and together they form a network of sensors. The data they collect is then sent back to much larger computers for analysis, all to help with climate research.

Sounds simple? Just throw a few small wireless computers into a field and leave the rest to the computers and the wireless network? Well, no – there’s a huge number of science and engineering challenges involved.

Sleepy sensors

One of the project team’s main challenges is power consumption: these sensor computers, also known as sensor nodes, are usually placed in hard-to-reach areas, like mountain slopes or hazardous chemical environments. This means that the period of time between replacing the batteries in the sensor nodes must be as long as possible.

The usual approach to maximizing battery life in sensor nodes is to turn off the radio for much of the time, but communication between a pair of nodes requires that they both be awake at the same time. That means sleep periods must be carefully timed between nodes that wish to communicate with each other.

There is another problem though. The sensors’ radios are only short-range, so for nodes to communicate, messages usually have to be relayed by nodes in between. The timing of sleep periods may be quite complex in order for a message to be successfully transmitted. Messages that must experience multiple hops may end up being delayed while all of the nodes take a siesta.

Hop to it

To support such multihop communication, the sensor nodes have rules to determine the paths that messages must take to reach their destinations. The rules attempt to find the best path to use for each pair of nodes, minimising the number of hops, sharing the load across the network, and of course, finding the most energy efficient path between sensors dotted among the grass and rocks.

The project team hope to reduce the amount of power the radios use by about half, while keeping delays and dropped messages to a minimum. Since the radio accounts for most of a sensor node’s energy use, reducing the energy consumption by half means almost doubling the amount of work that can be done using the same size batteries.

If such improvements in energy consumption can be made on the Cumbrian hills, they can also be used in other mobile devices. How would you like your mobile phone standby time to be doubled? Not only would you get more time between charges, using less energy is good for the Earth.

Finally, back to the sheep. A tiny computer lying on the ground needs an aerial sticking up to improve its transmitting and receiving range for the same amount of power used. How would you design the aerial so a careless sheep doesn’t mistake it for a grass blade?