Managing hazardous slopes using resilient IoT sensors and real-time processing (slopeRIoT)

Posted on

< Project Overview >

Introduction

Landslides pose a risk to linear infrastructure such as roads. The project here was to integrate resilient IoT into our proven sensors to live stream key data to our post-processing workflows. The key project partner was Transport Scotland, and the consultancies working on their behalf (BEAR, Geo-Rope and Jacobs) with a developing partnership with Cumbria County Council that has been Covid disrupted. The project partners manage the road networks, the nature of the collaboration was for us to trial and implement new technologies to improve their ability to manage the network more safely.

Project aims

The primary aim was to rapidly share useful information to our stakeholders from our IoT sensor network to trigger actions to reduce risk. Known high risk-high value sites can be monitored intensively with high labour inputs, and/or, mitigated with hard engineering with costs measured in millions. Real-time and near real-time IoT sensor architectures offered promise in widening monitoring to network scales and allowing data informed decision making to keep cities linked and minimise threat to life. There are many IoT barriers to deploy such sensor networks, both technological and in integrating into existing decision making hierarchies. The primary barriers addressed in SlopeRiot were Business Case and Businesss processes: lack of understanding of the full landscape of possible architectures for a possible IoT solution; incorporation/streamlining of IoT based applications / decisions with existing business processes; lack of IoT adoption by other dependent organisations (e.g. supply chain) and internal process owners.

The follow-on part of the work has the aim to begin coding a web-based portal to host processing of data for the project partners, and, to investigate efficient data handling in the cloud. A minor objective was to continue to develop circuit board based IoT GPS tracking sensors, at a cost to be considered disposable for slope safety.

What was done?

Set up resilient streaming sensors; used live and near live processed data to identify and characterise slope failures and precursors to failure; evaluated streaming sensors, and suggested how our work could lead to changes in slope management. Cumbria County Council were a new key stakeholder who were not original project partners, a link that developed as a direct result of Pitch-In.

In the follow up a prototype portal, ingesting and processing rainfall data and threshold conditions was implemented, and, integration of the deformation tracking PIV code has been progressed substantially. the IoT GPS systems are awaiting deployment for full field testing in Cumbria, on behalf of Cumbria County Council.

Results

The project has achieved a permanent move by Transport Scotland to fund and integrate live streaming at a key risk hotspot, using our equipment and communication designs and workflows.

A number of sensors are still in place, streaming data, all of which is being used both strategically and operationally. We have also moved two proof of concept streaming instruments to the point of site testing, but, Covid restrictions have meant this has not been possible yet.

Deliverables and other tangible outputs

Publication (Pitch-In award ref is awaiting a change by MDPI as it was missed off at proof stage: Khan, M.W.; Dunning, S.; Bainbridge, R.; Martin, J.; Diaz-Moreno, A.; Torun, H.; Jin, N.; Woodward, J.; Lim, M. Low-Cost Automatic Slope Monitoring Using Vector Tracking Analyses on Live-Streamed Time-Lapse Imagery. Remote Sens. 2021, 13, 893. https://doi.org/10.3390/rs13050893

Publication (Pre-Print): Rupert Bainbridge, Michael Lim, Stuart Dunning, Mike Winter, Alejandro Diaz-Moreno, James Martin, Hamdi Torun, Bradley Sparkes, Muhammad Khan, Nanlin Jin The predictability of shallow landslides: lessons from a natural laboratory. https://doi.org/10.31223/X52W2R

Impact Case Study: The work has formed an integral part of a REF submitted Impact Case Study (can be viewed on request); the Pitch-In funds are the reason for the upscaling from Scotland to the Lake District via Cumbria County Council and the British Mountaineering Council

Impact

Pitch-In has led to a permanent change in monitoring using our equipment specifications and workflows at the A83, at the cost of Transport Scotland. It will (post-Covid) allow for further extension of our Pitch In derived methods to be used in Cumbria at a number of sites. The work has directly fed into a successful Scottish Roads Research Board award (£85k) in which a Newcastle PDRA will receive 3 months of salary. Work with/for Cumbria County Council using the Pitch In findings will be able to progress as Covid restrictions ease, at this point we begin to investigate the potential for spin-out or employment.

Next steps

As Covid disruption begins to settle the next step is an EPSRC Standard Grant bid to upscale and roll out our Pitch In findings to multiple sites, using the Pitch In partners as official partner organisations on the grant bid.
The web/cloud based portal is ready to deploy for rainfall operationally, and, a meeting is due in April to discuss integration into standard workflows, and, ways to progress the use of the portal.

Lessons learned

The project allowed to focus more on the communication of data from sensors to the workflows; it has been the key thing to drive adoption, as having live information is key to operational usage by stakeholders.

We were attempting to push on many workfronts at once, and in two locations. In hindsight, doing less, and doing to full completion would have been better, and, perhaps in one location.

The major barrier has been the lack of my time to drive the project. Without funded PI time and buy-out, it has been hard to strategically engage at key times.

What has Pitch-In done for you?

Pitch-In has allowed us to explore, develop and then verify the benefits of resilient data streaming to manage landslide risk in real time, or near real time. It has opened doors to new ways of working with data and made planned grant bids more competitive.

Project lead:

Dr Stuart Dunning – Newcastle University