Image: Oliver Sved/Dreamstime.com

Analysis

Bringing sewer inspection and rehabilitation planning into the digital age

5 November 2019

Andy Taylor, chief engineer at Atkins, outlines how AI and automation can radically transform approaches to drainage inspections and rehabilitation.

Ageing infrastructure, doing more for less, greater public expectations, increased regulator scrutiny on performance, pressure to embrace new technology – water companies, local authorities and transportation operators face these challenges, and more, every day. But could we take one of these challenges – new technology – and change it into a solution rather than a problem? 

Every year, the lives of people across the UK are ruined by raw sewage flooding their homes due to poorly maintained sewers. The solution to this could be as simple as using technology in the right way to help us better understand and maintain our drainage assets.

A big problem, a digital solution

At its heart, asset management is about understanding what you have, its condition and prioritising the risk of failure to allow proactive planned maintenance. Gravity drainage infrastructure is no different, but unfortunately sewer inspection and rehabilitation planning have largely remained unchanged over the last 20 years.

CCTV surveys are conducted on site, causing disruption to local communities and generating survey data of variable quality for a resource-constrained client who is unable to review, interrogate and convert the vast amounts of data collected into useful information.

So how can we change this so that our assets and our communities are protected? At Atkins we have begun applying AI technology – previously used in our transportation and tunnel projects – to sewers, to automatically identify defects. This will improve data quality, reduce time on site and, by removing human intervention, reduce costs.

We recognise though that AI only solves the problems of data collection and quality – the client is still faced with masses of data ripe for automated processing. This is where a new tool developed by Atkins comes in.

By using RATS, CPS was able to quickly and accurately scope the remedial work required to allow a schedule of rehabilitation to be passed to the contractor

RATS (Rehabilitation Automation Tools for Sewerage/Stormwater) automatically reviews CCTV survey data, coded in accordance with the Manual of Sewer Condition Classification, and recommends appropriate rehabilitation schemes based on critical modes of structural failure. It is built up from a series of logic statements which have been compiled into a decision tree and programmed for automation.

These rules determine whether the pipe requires isolated repairs or manhole-to-manhole remediation is to be undertaken. For complex combinations of defects, the tool flags where an “engineer review” is required.

RATS: automation in practice

As part of the testing of RATS, Atkins worked with Connect Plus Services to review the CCTV survey data associated with 22km of highway drainage on the M11, with RATS used to automatically process large amounts of HADDMS (Highways Agency Drainage Data Management System) CCTV data.

RATS was chosen to undertake the work to decrease the time needed to understand what improvements were urgently required for investment planning.

The HADDMS CCTV data was processed through RATS and the results reviewed. This concluded that for the 288 pipes, 230 did not require any remedial work. Of the remaining 58 pipes, 36 required patch repairs, three required excavation and one pipe needed to be relined. A total of 18 pipes were identified for engineer review, these included physical assessments on the extent of damage or complexity of defects.

On this trial we were able to demonstrate that we could automate elements of our drainage design process, for example the pipe network’s structural and surface defects, prioritise assets as well as recommend potential renewal treatments.– Quintin Viljoen, Connect Plus Services

In addition to recommending sewer rehabilitation improvements, RATS was used to identify operational problems recorded in the CCTV data which may be responsible for blockages which could then cause flooding. This highlighted areas of root intrusion, debris, encrustation and obstructions for which planned maintenance operations could be scheduled. In addition, a few anomalies in the CCTV data were flagged by RATS for investigation.

This work could then be prioritised against the severity and consequence of failure to allow an optimised programme to be developed and undertaken during planned lane closures, to avoid reactive work.

By using RATS, CPS was able to quickly and accurately scope the remedial work required to allow a schedule of rehabilitation to be passed to the contractor. The cost was approximately 50% less than the cost to undertake the work manually and halved the programme to less than a month, thus allowing improvements to be implemented quicker.

Quintin Viljoen, professional head linear, at Connect Plus Services, commented: “We have trialled the use of RATS on the M11 to understand the potential of this tool in our drive for innovation, efficiency and digital asset management. On this trial we were able to demonstrate that we could automate elements of our drainage design process, for example the pipe network’s structural and surface defects, prioritise assets as well as recommend potential renewal treatments.

“The recommendations made by RATS concurred reasonably well with our own engineering recommendations. The use of this tool has the potential to provide notable design efficiencies.”

With all results generated in GIS, the output from RATS can be linked to other datasets such as land use, traffic flows and topography to further refine the output for other organisations such as water companies and allow outline scheme briefs to be generated semi-automatically.

Could automation transform your projects?

This new digital age provides significant opportunities to better utilise the vast quantities of data generated from CCTV surveys to generate savings and increase value for clients, while creating better outcomes for end-users.

With a number of companies investigating the use of AI for defect recognition in gravity pipes, the collection of accurate defect information is not far away. When combined with tools such as RATS, there is great potential to streamline the process from data collection through to automatic sewer rehabilitation scheme development. As such 2020 could be the year we see a radical transformation of how we manage our precious drainage assets more efficiently.

For further information on RATS please contact ratscontact@atkinsglobal.com

Image: Oliver Sved/Dreamstime.com