Hiding In Plain Sight: Improving asset efficiency from existing data
Sweco author: Mark Narbrough, Technical Director (Asset Management)
Our water consultancy clients’ most pressing requirements are typically along the lines of “We want to use data to reduce the number of service failures affecting our customers” and “we want to use data to minimise the impact of failure through proactive intervention”. My response is usually “You can do both – you already have the data to make a significant impact on these problems…and it’s in plain sight”. Furthermore, I add that it is often easier and cheaper to analyse and implement than they might think.
This article looks at what information in fact is ‘in plain sight’ and how to reap the benefits of it. For clarity, I am focusing here only on rotating plant (things with motors) but the same principles apply to many assets across water infrastructure.
When my car develops a fault, a lamp lights up on my dashboard…
The most critical faults have their own lamps and warn me to stop the car. The less critical things generate a common warning signal advising me to take my car to a specialist who can diagnose the fault.
That warning prompts pre-emptive action that, if addressed, helps to avoid a future critical failure.
Similarly, in the water companies when there is a critical fault onsite, an alarm occurs in the central control room and a work order is raised for diagnosis and resolution, however, for many that is where the data journey ends.
What I find is that there are often no equivalents of the common warning lamps for water infrastructure assets and hence no one is sent to diagnose and resolve the fault. The other common issue here is that the volume of issues that are deemed ‘critical’ faults is so great, that even they cannot be attended to when they do come to light – but that is a separate issue for another article.
Technology & data
Some cars have a service light, a lamp that tells you when the service is due, this is typically based upon the distance travelled and time. This is one piece of information (metric) provided to the driver, who is typically also the ‘maintenance manager’, derived from collected data.
In the water utility, this development of metrics from existing data is the quickest and cheapest way to gain insight into asset performance and to address the early intervention or reduced service failure ambitions.
For rotating plant, the key two pieces of data are the failed and the running signal. From these, there are a vast array of metrics that can be gleaned, but the most valuable ones are those on the right of our diagram above. They shorten investigation time, help diagnose the root cause and help answer the key questions:
- Go or don’t go?
- If go, when to go?
- When I go, what skills are needed?
A wealth of information therefore can be accessed and used meaningfully from only two signals. When equipment is combined in a set such as duty standby pairs, as they so often are, even more intelligence starts to land in our laps.
The use of this available data to generate service lights (information that can anticipate problems before they become failures) is rarely seen. The sector regulator is encouraging better use of existing data, perhaps this is it.
People, process & plant
The people side of these projects fascinates me, especially when I hear about the type of initiative that “didn’t work”. When I dig deeper, the things I tend to find are that these projects are often technology-led and as such:
- There is no root cause process
- The wrong metrics were being derived
- If the metrics are correct, they are displayed to the wrong people
For successful data implementation, there are five key elements:
1. Have the specialists, who can undertake the root cause and diagnose the faults. As per the car analogy, your seasoned mechanic in the garage. This person can answer the questions:
– Go or don’t go? – “I’ve reset that for you come back if it comes on again”
– When to go? – “You need a new widget now”, “you need a new widget soon”, “you can get second-hand widget”
– What skills ?– “You can get it from Halfords”, “we can do this for you, it’s specialist”
2. Have a feedback process in place to continuously improve the root cause analysis
3. Be able to export data from the data acquisition systems and contrary to popular belief this function does not need to be real time to deliver great value (start with running and failed data for the rotating plant)
4. Have tools for analysing the data and calculating the right metrics, this can be as simple as a spreadsheet without a fancy dashboard display, to business intelligence tools and data lakes – my experience is new solutions tend not to be required to implement this approach
5. Have formalised processes for responding to faults or failures and routine or planned maintenance so business implementation is swift and efficient
When the above are ALL achieved the advantages and positive outcomes can be significant.
For one of our clients, tangible benefits were realised in the reduction to reputational damage and extended asset life, with a conservatively estimated saving of over £250k per annum in fines and recovery costs. This makes me confident that this approach leads to a more sustainable, lower whole life cost of operation. Something I strive for daily, whether it’s for the smooth running of my car or my customers’ projects.
My findings are that, in order to expose the ‘service indicators’ from your data, there is usually no need for new software, or new hardware. Uncovering the information is the easiest part to achieve, but to realise the benefits, it is good process and great people that will make it succeed.