Wide disruption is already occurring within industries ranging from transportation to garden maintenance as a result of Artificial Intelligence (AI) and Robotics. Automatous vehicles are expected to remove the need for taxi and lorry drivers. These impacts are affecting resale values of previously coveted taxi licenses which once traded at up to $1 million. The water industry isn’t immune to this trend.
Widespread automation has already occurred in countries such as The Netherlands, where remote operation of treatment works is the norm. Many of the local councils in Holland have centralised their site operations which then enable remote operation. This focus on efficiency has mainly been driven by requirements to reduce customer bills by 2 per cent per annum.
In the UK, the risks of job automation, as identified by PwC’s March 2017 UK Economic Outlook report, appear highest (based on employee numbers) in transportation and storage (56 per cent), manufacturing (46 per cent) and wholesale and retail (44 per cent) but lower in sectors like health and social work (17 per cent).
Although the UK water, sewage and waste industry only employs 0.6 per cent of total employees, the industry has the highest potential 62.6 per cent risk of automation, according to PwC. A typical water utility is made up of operations, capital delivery, asset management and customer service. PwC predicts that for those with GCSE-level education or lower, the estimated potential risk of automation is as high as 46 per cent in the UK. This falls to only around 12 per cent for those with undergraduate degrees or higher. This would suggest that operational staff are at most risk since many have an education below GSCE level.
However, if we look closer at examples of AI deployed in the water industry, the story of its potential impact on job security is more complex than PwC’s analysis suggests.
The closest we have to real AI in action in treatment works and networks are Real Time Control (RTC) systems created by companies such as Hach Lange and Veolia. Both companies have used RTC in a range of applications. For example, in activated sludge, the systems work in similar ways using complex algorithms and in situ instrumentation to determine the ideal Dissolved Oxygen (DO) concentration for any given load. By calculating incoming ammonium load through monitoring ammonium and flow, it is possible to calculate the required DO set point for any given treatment condition. These set points can then be delivered to the blower control system to ensure robust and efficient plant operation.
The value of these automated systems is clear: it has saved up to 25 per cent electricity use and up to 50 per cent chemical use without major human input. However, these systems still rely on operational input such as cleaning weirs and maintaining equipment.
Nonetheless, there is an impact on the requirement to constantly optimise the works. In this example, it is process scientists and optimisation specialists that are most at risk from AI. So, contrary to conventional wisdom, it is the more skilled workers who are more likely to face a negative impact from the rise of AI.
Slowly but surely, companies such as Veolia are connecting all aspects of treatment works, networks, rainfall data and weather as shown in figure 2. As all these connections are made, real data will be more readily available for customers and could be used to change customer behaviour. However, as over 90 per cent of wastewater treatment works are very simple low tech percolating filters, it is difficult to see AI having a major impact here.
AI is highly unlikely to impact the water industry in the short term. In the longer term, as the industry moves away from Enhanced Capital schemes, and pressure to reduce customer bills increases, AI will become more prevalent.
There appears to be some way to go before AI starts to make a significant impact in the water industry, but the landscape continues to change quickly. Intriguingly, and perhaps counter-intuitively, it seems AI will impact process scientists and optimisation specialists more than those engaged in operational activities such as clearing blockages and maintaining equipment. These will still be crucial activities which AI can’t replace, meaning the people doing this work remain critical to delivery.