Landslide detector wins enterprise award

An early-warning device for detecting landslides has won a Loughborough University enterprise award.

The invention, developed by researchers in Loughborough’s civil and building engineering department, uses real-time acoustic sensors to listen for movement in the soil.

Following work funded by the EPSRC and built in collaboration with the British Geological Survey, researchers last year conducted successful trials of the sensor and won the commercialisation category of the university’s Enterprise Awards 2010.

Existing monitoring systems measure the physical movement of devices planted in the ground and are checked at regular intervals to provide an indication of the average size and speed of movement.

Loughborough’s device, called Assessment of Landslides using Acoustic Real-time Monitoring Systems (ALARMS), detects high-frequency stress waves produced by soil movement. Because they are continuously monitored by wireless communication, they can be used to calculate soil movement in real time.

The challenge was finding a way to detect the very small stress waves given off by soil movement, said Neil Dixon, professor of geotechnical engineering at Loughborough and co-inventor of ALARMS together with Matthew Spriggs.

‘When rocks fracture they give off lots of energy as acoustic emissions,’ added Dixon. ‘But most landslides are in soils and you get very low energy when two soil particles move against each other and the stress wave loses energy very quickly in the ground.’

ALARMS uses a steel tube called a wave guide to conduct the signals out of the ground. This is placed in a borehole in the ground filled with sand or gravel that produces more energy when moved than the soil itself, making the signal easier to detect.

The device also detects high-frequency waves of around 20-30kHz because lower frequencies within hearing range, although they lose less energy and are easier to detect, would include background noise from other sources.

Dixon carried out tests to determine how the acoustic signal produced by the gravel related to the speed and size of movement in the ground, regardless of what kind of soil it is.

‘Because we used the same backfill and wave guide every time we can actually calibrate it,’ he said. ‘If we get a certain amount of acoustics, we can relate that to a displacement rate.’

Measuring such high-frequency waves requires very fast computing power, so Dixon, in collaboration with the British Geological Survey, also developed a low-cost instrument to process and transmit the landslide data from the wave guide.

‘It counts the number of times the signal goes above a threshold every 15 minutes and it’s that number we relate to our calibrations to tell us how fast the slope is moving,’ said Dixon.

‘This has only been possible because of the development of printed circuit boards and the processing that can now be put on small chips. We couldn’t have done this five years ago.’

Dixon has plans conduct a second set of trials this winter on infrastructure slopes around roads and railways. He also hopes for further EPSRC funds to allow him to redesign the sensor to minimise power and cost.