Last updated on May 17, 2018 by Dotsquares
In an interesting interview with Yotta’s chief product and technology officer, Manish Jethwa, ZDNet unveiled how data is helping maintain the roadway’s infrastructure, and how with IoT sensors and machine learning this process is going to become even more efficient in the near future.
Yotta is a tech agency working in the field of infrastructure asset management, by using data from sensors and surveys. For the last 25 years, as Jethwa says, the organisation has been working on the infrastructure of highways by providing local authorities and national agencies like Highways England with useful data.
Jethwa says the required data is collected through the means of surveys and sensors. For this purpose, the company also builds its own survey vehicles that go around the highways, collecting detailed information of the road network. He also mentioned using laser technologies to get information about every millimeter of the road to make their detailed models.
This data then helps the responsible authorities to make a strategic decision on where the treatment is required and how they should invest in it.
The Alloy Application
Over time, as the agency accumulated more and more data over the assets used for the roads infrastructure including streetlamps etc., they started investing in it for the creation of an automated asset management system, which they call the Alloy.
The company has been working on Alloy for the last couple of years. They are using it to converge huge amounts of data like the information of every nanometre of the assessed infrastructure into meaningful insights that the local authorities can use to make more effective decisions. Such information enables local authorities to see and solve the issues that need immediate attention.
The obvious benefits of the software have encouraged the company to become more involved with the technological aspect of their jobs and use the more advanced technology to make the data more useful for improved asset management.
Introduction of IoT
Jethwa points out that the lack of demand and consequently the production of survey vehicles has kept their prices very high, around half a million pounds. The expensive survey, therefore, was only organised annually at its most frequent.
The life cycle of these assets is also a major concern that needs to be resolved, and this is where IoT comes into the picture. Jethwa proposes using IoT sensors in the assets that are nearing their end, as most of their depreciation is based on the sensors that are already embedded in them. Thus replacing traditional sensors with IoT devices will not only prolong their lifespan by a couple of decades, but will also make them more effective in terms of collecting more useful, and real-time data.
Future with Machine Learning
The interview also brought light to the inter-dependencies of various assets and the other factors like weather that can impact the workflow of infrastructure maintenance. Using an example of drain network, Jethwa illustrates, that when gullies around the roads get blocked, they greatly affect the surrounding infrastructure as well. Previously, such issues were being addressed by using a single map that showed all the assets on it.
Now, with the help of advanced technologies, a multi layer map can be created to help the connected asset-management. Such a database will bring to light the varied inter dependencies between different assets. With real-time insights from IoT, appropriate actions can be taken to optimise the use of resources before a minor issue can become a major problem for the surrounding assets.
All this optimisation, Jethwa says, becomes more important when we consider the growing stress over the infrastructure which is down to the increase in population, traffic, and weather anomalies. With time, it is becoming more and more apparent that all these factors are inter-dependent, but understanding this on algorithmic levels is complicated, therefore, a task that can be more easily accomplished with the use of machine learning to help out.