Interview with Takashi Kato: How Fracta Uses Machine Learning & Big Data to Optimize Infrastructure

Published on by in Technology

Interview with Takashi Kato: How Fracta Uses Machine Learning & Big Data to Optimize Infrastructure

Boldly Go Where No Man Has Gone Before: Achieving Your Dreams By Driving Water Infrastructure Efficiency and Affordability through A.I. and Machine Learning

Ahead of Fracta joining us at the World Water-Tech North America summit , we spoke to Co-Founder & CEO Takashi Kato to learn more about how Fracta uses machine learning and big data to optimize infrastructure investments – starting with a water main pipe replacement prioritization solution for U.S. water utilities.

We live in exciting times with the digital transformation occurring around us where our decision making is enhanced through A.I. and we can visualize the aggregate results of thousands of data variables with millions of processing cycles - Something our human minds and spreadsheets could not compute.

When we were young, we were each captured by science fiction and our dreams of the future included the belief that our wisdom and understanding of the world around us would be enhanced by computers and robots. From Star Trek to Japanese anime our imaginations have been captured and have influenced our expectations of what we could achieve in the future. These dreams fueled my passion for education and my technology entrepreneurial focus and drive.

What we did not realize was that the infrastructure of our modern society would age and create financial and public health challenges and that our technological advancements in computing sciences with powerful algorithms would find a practical application of first saving and protecting our standard of living on earth.

As an example, our historical engineering efforts built our life sustaining water systems. Our maintenance procedures have helped avoid reactive or costly emergency failures by moving towards planned activities which provide a 12-18% cost savings benefit. However, the ability to further mature infrastructure asset management best practices has required A.I. and Machine Learning to successfully move from planned to predictive analytics to gain an additional 20-30% in cost efficiencies. This digital transformative evidence has been documented in several case studies on accuracy and cost savings with our underground water pipes using the Fracta Machine Learning solution.

 

Case Study: Large-Sized Water Utility

Five years of water main break data from a large utility with 3,395 miles of pipe was used to compare how each model would predict the actual pipe’s failures. To do this, part of the data set was withheld from the machine learning model to demonstrate the accuracy of its predictability.

The Fracta Machine Learning model captured 26.2 percent of the historical pipe breaks as part of its analysis of the highest risk or worst 5 percent of pipes that are predicted to fail. This 5 percent of the 3,395 miles of pipe identifies 139.5 miles of pipe as the highest risk pipes that are predicted to fail.

The age-based model captured 26.2 percent of the historical pipe breaks by identifying the worst 7 percent of the pipes. This 7 percent of the 3,395 miles of pipe suggests that 195.4 miles of pipe would need to be replaced to avoid the historical breaks.

In comparing the two models, the machine learning model was 28.5 percent (2 percent/7 percent) more effective in identifying pipe breaks over the age-based model.

Fracta is a US-based company, born as a technological venture in the Silicon Valley of California. Fracta utilizes A.I., notably Machine Learning to predict the remaining useful life and therefore the optimal replacement time of water main pipes using pipe data such as age, material, install year and break history combined with over a thousand of additional environmental variables such as soil, weather and population specific to each pipe segment. The predictive Fracta Machine Learning solution is being used in over 20 U.S. states by engineering firms, utility managers, and water distribution systems professionals to leverage their condition assessment operating budgets for direct pipe inspections and leak detection and maximize their capital expenditure on renewal and replacement of their highest likelihood of failure and consequence of failure water mains. (Discover more)

This machine learning technological advancement of creating a more accurate digital desktop risk-based (not subjective) model of the entire water distribution system at a fraction of the cost of traditional methodologies is truly a digital transformation milestone for the U.S. water industry.

 

Reduce Costs with an A.I. Machine Learning Digital Desktop Condition Assessment

The Fracta Machine Learning solution for water has overcome our limits on our human ability to gather, process and understand all the information needed to optimize a decision.

We are visual beings and 90% of the of the information is absorbed by the brain as visual. We process visuals 60,000 times faster than text. As a result, Fracta has partnered with Esri ArcGIS, the leading GIS firm for utilities and governments to create a data intelligence visual experience for expedited risk mitigation and decision making for field crews and utility management. This platform allows for the additional integration of other water related data and multi-sector infrastructure asset management planning.

 

Visualize and Apply Results to Support Asset Management Decision Making

Utilities can now focus on asset management and risk mitigation strategies. Engineers, planners, and management can now make quick, accurate and affordable decisions regarding:

In 2018, Fracta got acquired by the multi-national corporation in the water industry (Kurita Water Industries Ltd.) and received a significant boost into the international water market.

The goals and objectives of affordability, public health, customer service, and sustainability are also the international drivers of the acceptance of A.I. and Machine Learning. In Europe, especially in developed countries, water main pipes are aging and reaching their end of useful life, which leads to increasing demand for efficient replacement, maintenance and a reduction of leaks of water mains. By 2019, Fracta in conjunction with Marubeni began to develop a water main condition assessment software which predicts the likelihood of water main failure in the United Kingdom with the Northumbrian Water Group, in the North East Region of England, UK.

Fracta’s proprietary algorithm to predict the condition of linear assts like pipelines has also expanded to gas pipelines belonging to Toho Gas. The Artificial Intelligence (AI) powered algorithm will utilize over 1,000 environmental variables to assess the condition of the gas pipelines, the first undertaking of such in the world.

The future of A.I. and Machine Learning is exciting and will continue to expand our knowledge and understanding of our surroundings like using a cloud-based SaaS application mobile device “Trekie Tricorder” to quickly analyze a thousand details of a new planet to explore. I believe that the intelligence we gain from A.I. and Machine Learning will help us become better citizens and stewards of our natural and built environments.

I look forward to gaining additional insight at the World Water-Tech North America summit of how Fracta‘s drive and passion can be combined with other technological advances and dreams to create a more affordable, safe, resilient and sustainable future.

The interview was originally published by World Water tech North America

 

Doug Hatler Chief Revenue Officer, Fracta will be joining Takashi at the summit. Doug will be speaking on the panel, From Machine Learning to Smart Water Grids: Driving Resiliency Through Digital  at 11 am on October 30. He will be joined by

Session Chair:
Samuel Saintonge Principal,   XPV WATER PARTNERS
Speakers:
Rebekah Eggers WW Leader – IoT for Energy, Environment, & Utilities IBM
Cecil McMaster,  CIO NYC DEPARTMENT OF ENVIRONMENTAL PROTECTION
Albert Cho VP & General Manager, Advanced Infrastructure Analytics,  XYLEM
Ting Lu Business Practice Leader – Digital Solutions,   CLEAN WATER SERVICES

Media

Taxonomy