Ashay Prabhu talks to SCW about how advances in data processing and sensors has made it possible to predict the future and save organisations significant amounts when it comes to their assets
Ashay Prabhu has made a great deal of money predicting the future. Ironically, however, when he started Assetic 15 years ago with co-founder Joel Brakey, the duo quite hadn’t foreseen the path ahead.
Advancements in technologies such as data collection, processing and analytics have impacted and shaped the company’s service offerings, and evolved the technology they created.
Global asset management software and services company Assetic is based in Melbourne, Australia. It provides engineering and financed based asset management professionals with a central repository for all their assets data. By doing this, Assetic enables organisation to make better long-term decisions, thanks to unique algorithms and degradation profiles that forecast future capital expenditure.
But it didn’t quite start that way. The company started out as an analytics company that looked at data from local government clients. “Eighty-five per cent of the world’s infrastructure is with local government, and there is no bigger data than local government,” says Prabhu.
However, an asset is an asset, and in the last five years the company has branched out into asset management for a wide range of other industries such as water, rail, energy, ports, mining and facilities management including higher education, community housing, government-managed facilities and stadiums.
“Whether you look at road, rail, transport tunnels, bridges, buildings, hospitals, schools etc, all the assets have genetics, and there are certain patterns within infrastructure that can be extrapolated from the genetics,” he says.
“We started to realise that once you started to model these genetics and look at emerging patterns you could actually see the future city right now. You could see the future transport agency or the future school portfolio now, and you could see four, five, six or even seven versions of these depending on how you made your decisions.
“You can apply genetic science to big data, and model it out to see what the future may look like if something different was done to it. For example, if I cut my budget, what would happen? “If I want to have healthy assets that are heavily utilised, how much money would I need? How much extra funding would I need to change the way that I give services to my citizens? I might want to look at congested roads -- shall I put freight back onto rail? I could model it and see the impact so that politicians can see the future now, and make decisions based on that,” he says.
In the last two years, assets themselves have started to become smarter. Instead of going out to collect data, asserts are using sensors to transmit information back to an organisation.
Making assets smart is vital to a smart city. Rather than just alerting citizens to a situation as soon as it occurs, sensors deployed to assets provide a pre-emptive facility as constant and near real-time monitoring anticipates problems way in advance therefore ensuring that a critical situation never arises in the first place.
“Essentially,” says Prabhu, “It is like putting steroids into the system because the technology that we have, which is analytics around big data just keeps getting better and better. In the old days to run an analysis of a small city with 15,000 assets over a 20-year prediction period would take us 48 hours. Now it takes under two minutes.
“Twenty years ago most organisations no matter what the industry, the data they had on a spread sheet was no more than 15 columns because that’s all they could collect. The data now, even from the most immature organisations is about 200 columns. So when you combine 200 columns with one million assets, and people want to look 50 years into the future, the permutations and the computations are more than the number of atoms in the universe!”
Predicting the future from the present, with the help of historical and real-time data which is mathematically optimised, is saving organisations from the smallest city to the largest rail company and everything in between, a great deal of money.
Last year Assetic carried out futuristic modelling scenarios for South African Rail, and demonstrated three different scenarios of how it could cut funding by 30 per cent and achieve an improved network in a decade.
“We have carried out similar work with a lot of local governments in Canada, local governments and water authorities in the US, and over 100 city councils in Australia, and every single use case, they have cut their future infrastructure gap by massive proportions from 20 to 40 per cent without spending more money,” says Prabhu.
Take for example the experience of Corangamite Shire Council, on the South Coast of Victoria, Australia. It manages $800 million of vital community infrastructure. As an early adopter of strategic asset management methodology, Corangamite is now seeing real benefits in reduced asset consumption across its entire asset portfolio.
Over a period of 10 years, Corangamite has applied the outcomes of Assetic to improve asset performance and condition, whilst also reducing the rate of asset degradation and the associated capital renewal and maintenance backlogs. Importantly, Corangamite Shire has been able to do this without significantly increasing spending; rather, it used Assetic’s prediction and capital prioritisation model to determine better ways of spending existing funds. Through this more-informed decision-making, Corangamite is able to spend available funding in the most optimal way and reduce overall asset consumption rates.
Brooke Love director works and services, Corangamite Shire Council says, “Assetic aids in predicting intervention for renewal works, enables a program of works to be developed, quantifies the funding required to ultimately reduce the renewal gap or risk of renewal gap expanding."
According to Love, the change of approach enables Corangamite to optimise service level outcomes and capital and maintenance expenditure. “If you spend small amounts regularly, as recommended by the outputs of the Assetic optimisation process, then it saves you from the big spend in the long run,” she says.
In the City of Windsor, Ontario, Canada, municipal managers were grappling with this question: How to deliver an optimal level of service while utilising prudent capital financial planning principles and at an acceptable level of risk.
It decided to try Assetic software as part of a pilot Strategic Asset Management (SAM) programme. The council wanted to see how assets would fare looking at a range of timelines and funding scenarios. Assetic provided the means by which a range of options could be modelled. Management gathered data and assembled a working group from GIS, public works/operations, asset planning and finance, risk management and planning, as well as executive leadership.
They utilised objective condition data and knowledge contained within their CMMS network, identified key intervention strategies based on desired levels of service and condition, and rapidly modelled multiple funding scenarios. In five short months, senior management presented to council three different funding scenarios over a 20-year timeframe. Since then administration has developed 20-year projections for several other asset classes and is currently developing long-term facilities models, all of which were carried out in a single year.
The City of Windsor gained valuable insight and developed a framework for implementing sound infrastructure investment strategies. In addition, managers got a 20-year asset renewal investment strategy, charts showing the resulting overall conditions, the impact on taxpayers for various service levels, and a detailed capital expenditure programme. This helped it dramatically reduce the time it took to generate and present the data, and helped to inform future infrastructure decisions.
The council discovered that a 0.5 per cent tax levy applied for 20 years would allow the city to deliver a more sustainable level of service and risk tolerance when compared to other funding scenarios studied. Using Assetic, SAM allowed management to deliver actionable information to the council about their options in terms of spending taxpayer money on roads. The results were so impactful that they led to requests for further studies for additional infrastructure assets.
Looking forward, next on the horizon for Assetic is artificial intelligence (AI): the mimicking behaviour patterns. “It is about saying, I want to look at 20,000 datasets for a pattern of behaviour that might be for heavy trucks running along the California highway and what behaviour does the road display,” says Prabhu.
“You can look at 20,000 datasets and create an artificial intelligence into the system, so if in the future, anything anywhere in the world mimics the California highway which is 10,000 trucks a day, travelling over 100km an hour with an Asphalt surface thickness of 500mm, then bang, without even thinking, the system will have a genetic pattern that mimics that behaviour,” he says.
Already, Assetic has partnered with an electronic road data capturing company in Australia. It has over four billion data points in its repository and has developed a system where you capture a picture of the pavement outside your house, onto your computer, turn on the AI module, and it will run through all available data to interpret the condition of that section of road. You will then get a generic rating, because you are comparing your image with thousands of other road sections, topography and terrain locality.
This use of imagery, says Prabhu, is adding to the overall power of analytics. While analytics is Assetic’s religion, it is totally indifferent to the sensor technology it works with. All its mathematical, analytical processing has been developed in-house, and while it has been taking AI patterns from third parties, it is developing its own in-house AI platform to propel its data analytics ever further.
Says Prabhu: “Technically we are agnostic to what type of sensors we work with, it could be a human being going out to monitor something, a few local sensors, or a high-quality sensor. What we have are all the data structures, all the hierarchies and all the algorithms, thousands of asset types, and as long as the data can come in, our system processor is ready.”
If you enjoyed this, you may wish to view the following:
Increased intelligence for assets
Tego’s innovative asset intelligence platform now supports major mobile OS and RF gateway protocols