How we are transforming the water industry with AI

From advanced metering infrastructure, to digital twins enabling network simulations, to the deployment of sensors at scale, the water sector has pioneered innovative data collection methods for decades.  But, as industry has begun to leverage that asset information to successfully invest in its networks and develop solutions such as preventative maintenance schedules, it has only just begun to realise that a similar dividend can be achieved through investment in AI solutions for its field workforces.

“2023 is the year of Artificial Intelligence and the time is now for water utilities to embrace this next wave of digitisation for their field workforces. Together with readily available global 5G, remote workforces have more ‘connection potential’ than ever,” says Shelley Copsey, CEO at FYLD. “Paper style solutions – either traditional paper forms, or digital forms with extensive drop downs and pre-formatted answers, are past their time.” 

FYLD is powering this AI revolution for the optimisation of field workforces, using AI to achieve outcomes for customers such as:

  • An 8-12% field force productivity uplift in as little as 6 weeks;
  • 35% leakage reduction through process redesign improving alignment of client and supply chain management;
  • 75% reduction in aborted jobs, for as little as 8% of traditional cost to stop leakage; 
  • Deployment of real time, high risk pollution event models; and
  • A reduction in field worker incidents and injuries of 20%, through a 30% uplift in safety interventions enabled by increased remote visibility.

Here are three ways FYLD is using AI to empower water utility field workforces and achieve better business outcomes: 

Effective management of remote teams

Typical spans of control in the water industry are upwards of one remote manager to six crews in the field.  These managers struggle to determine which crew and which issues need their attention, and figuring out what to prioritise is near impossible with traditional systems due to the lack of real time site visibility.

FYLD’s AI solves this challenge through transmitting site data (videos, voice, pictures and messages) to command centres in real time, which builds up a rich picture of the day’s operations, overlaid with AI-powered recommendations on where and what managers should prioritise.

FYLD’s unique state of the art Command Centre enables a laser focus for managers to see in real time opportunities to make job site interventions before problems emerge – reducing aborted jobs.

“FYLD’s AI also shines a spotlight on emerging workforce trends, creating actionable insights for operational managers so they can focus on taking immediate real time action needed to mitigate potential risks,” says Copsey.

As an example, FYLD can identify workers that are taking longer than dynamically benchmarked job durations to complete a piece of work and suggest reasons why.  Such as, are they frequently working on a part of the network where benchmarks need to be adjusted (for example, a known difficult valve, or an older network segment), is this systemic from a depot where trucks are not being filled with the right equipment to deliver a day’s operations smoothly, or is some training required?  

Supply chain oversight and real time permissioning

“Outsourced supply chains can make or break water utility operations.  Major water utility company operations are often outsourced to their specialist supply chain partners, with this level of mutual dependence meaning the digital systems used are crucial in enabling successful and sustainable service to customers,” says Copsey. “Research at FYLD has shown that on average standing time for the fieldworker sector is upwards of 20% on an average day.”  

Examples of reasons include waiting for access gates to be opened, specialist equipment, machinery or vehicles to arrive on site, through to simple things like not enough boards and barriers to safely barricade deep excavations.  This impedes the way a planned or reactive job is planned to be undertaken and crews either wait on site for long periods of time – often with a leak underway – for permission for a new way to complete the job, or abort it completely.

Water utilities are also often unsure who is on their site, the quality of work performed, and if safety protocols are being followed. Through seamless connectivity from site survey, to job plan, to supply chain outsourcing, FYLD’s AI helps build rich site data meaning jobs are planned effectively and briefed well to contractors. When changes need to be made on-site, permission is requested via the FYLD Command Centre, where operational staff can approve changes in real time.

  • A leading UK water utility implemented this FYLD workflow with their main Tier 1 contractor and reduced the time to fix leaks by 35%. This cost 85% less than industry norms, and is forecast to save over 450m litres of water in the first year of deployment alone.

Water utilities can also better understand who is on site, what they’re doing, and the condition of the asset when they leave by analysing info transmitted through FYLD, including video and picture metadata, and geospatial information from users and logins.

  • On a mains replacement project, a FYLD customer achieved a 12% productivity uplift in two weeks through subcontractors being on-site when their contract required them to be – through simple FYLD metadata analysis.
  • On a major US construction project, another FYLD customer increased site specific risk assessments compliance from 11% to 100% through data analytics, which unlocked real-time analysis of on-site workflows. This is converting into reduced insurance premiums, with a material reduction in incidents and injuries.

Generative AI for operational insights & field worker enablement

The sheer volume of operational activity that takes place in water utilities in any given month is tremendous.  The activity is also often unpredictable and highly reactive.

This also means it can be extremely complex and time consuming to produce operational summaries at month end that are (1) timely, (2) cost effective to produce, and (3) truly insightful and actionable.

“At FYLD, we are deploying Generative AI to help our customers tackle all of these challenges.  From a single job, to a region, to an entire network, we are working towards providing operational summaries that balance each of these often competing factors and enable our customers to take swift action to drive operational efficiencies, better network dependency and safe working practices,” says Copsey. “At the other end of the spectrum, surfacing the right piece of technical information to a field worker, at the right moment, is a key enabler of smooth operations and a reduction to standing time or aborted jobs.”

A variety of barriers have historically existed that have prevented utilities from being easily able to do this, including the digital-era problem of ‘an app for every problem’. This can lead to field workers stating confusion about which app they should access for what purpose.  And then there is the plethora of large process documents that any well governed utility has – that has also been incredibly difficult to surface to a field worker in a usable way.

“FYLD’s use of Generative AI through large language models is able to synthesise process documents and extract the key information a field worker needs in the moment to do their job effectively,” Copsey says. “And with FYLD being an end-to-end platform for the execution phase of field workforce activities there is no worker confusion about where to go to get this information – it is in the same place as their AI driven point of work risk assessment, their evidence/asset capture process, their job stoppages recording and their end of day work output records.”

Get in touch today to discover how you can transform your operations with AI, using FYLD’s revolutionary platform.