SmarterCX presents the Smarter Demos series, a 2-minutes-or-less look at some of the most innovative CX technologies and how they work.
In this 18th video in the Smarter Demos series, we take a look at some of the latest CX tech that predicts asset maintenance with machine learning. Renuka Uttarala, Senior Manager, Cloud Solution Hub, Oracle describes a solution that intelligently predicts asset maintenance for those in the infrastructure construction field. And Akintola Marke, Solution Engineer, Analytics, Oracle takes us through an example demonstration of a portfolio management company that uses predictive analytics to forecast damage costs and save the company repair costs.
Watch our interview or read the transcript below.
Renuka Uttarala: I’m so excited about the most recent solution my team built. It’s a big data solution, and it’s leveraging the machine learning capabilities. Machine learning is what it is all about, you predict, forecast the future and everything.
Though our solution, which we work for customers related to the infrastructure construction field and trying to solve the predictive maintenance, but this is open and works for anyone, like any customer, any mid-size, large, whoever the customer is.
Anyone who has a problem of large portfolio management, large property management, I want to predict the maintenance, I want to predict the risk. And all this risk mitigation, maintenance mitigation and all of those problematic areas, it works seamlessly.
Akintola Marke: Within this demo, we’re going to use a portfolio management company as an example. The company is managing a lot of commercial properties along the east coast. And with managing those properties, there’s been a lot of winter storms. And we’re using predictive analytics to predict how this weather will impact the HVAC systems of these buildings.
We forecast the damage cost that’s already been incurred during the winter season. We also highlight a few of the cities that have already had damages. We also highlight the predictability of these buildings having another occurrence of the systems breaking down.
In the next visualization, you see the average cost by day, and you see the average temperature by day. These essentially represent the costs spiking when there was a winter storm in the past.
In the top right visualization, we see the property system age. And then highlighted in red, we see the different buildings that were affected in the previous winter storm.
Towards the bottom, we see the buildings and the probability of these buildings being damaged. In this visualization, we have our predictive breakdown.
In our predictive breakdown, we get to see the estimated cost that could potentially happen with this future storm occurring, We can also see the different categories that the damage costs fit under.
With the middle visualization, you are able to see the probability of these different buildings being damaged and we can also see the geographic location of these buildings.
And in the actual breakdown, we’re able to see that only two buildings were affected from this storm. The temperature did drop below -5 degrees within the buildings, and the two buildings that were damaged, were in the northeast area.
The actual incurred cost was $1.2 million, which is a significant drop from the $7.5 million that we estimated.
In conclusion, our solution helped take preventative action with the damaged buildings. It helped us monitor the health of the actual buildings. It also helped us save cost maintenance and repair costs for the company as well.
To view the full Smarter Demos series, visit https://smartercx.com/smarterdemos.
Copyright © 2019 Oracle. All rights reserved.
This transcript may be edited for readability.