What is your technology?
VODA.ai provides machine learning (a subset of artificial intelligence) to predict water main failures in the next 12 months and over longer time frames for purposes of capital planning. VODA.ai’s machine learning engine is called daVinci which uses artificial intelligence to assess complex data sets specific to each utility and combines other data (e.g., satellite images, soil, land use, traffic, proximity to rail, etc.) to develop pipe break predictions for every pipe segment customized to each utility. VODA.ai delivers useful visual reports to justify investment plans and rate cases. The results are presented as shapefiles which are easily imported into GIS applications or via unique customer interfaces from VODA.ai.
What are the benefits to implementing your technology?
Pipe breaks and failures are becoming a greater risk for all utilities. A recent study by Utah State measured a 27% increase in pipe failures in just the past few years. VODA.ai’s daVinci engine helps utilities make smart decisions about which pipes to inspect, monitor, repair, replace or, importantly, to leave alone. The benefits include saving water, reducing costs of damages and overtime (breaks always occur at inconvenient times like nights and weekends), reducing non-revenue water, and avoiding premature replacement of pipes with remaining useful life. Some customers are using our risk-based rankings to test valves to ensure they will work in the event of a break. Some even use the rankings to help decide on pipe replacements when roads are scheduled for repaving. One utility used VODA.ai results to secure their largest-ever capital budget appropriation in order to address water main risks.
Has the technology been tested, demonstrated, or implemented anywhere to date?
VODA.ai’s technology is currently in use at many large and midsize utilities in the U.S. and abroad.
Customer testimonials include:
“In this challenging time, we are excited to use the predictive modeling from VODA.ai as another tool to help make better decisions in the future,” said Doa Meade, Director of Infrastructure Management for the LVVWD. “When the City of Las Vegas replaces pipes, we often have only two pieces of information to make expensive decisions: the age of the pipes and the history of pipe failures. With VODA.ai’s DaVinci machine learning toolset, we can now make more informed decisions.”
“The City of Tucson is working toward greater proactive management of our pipe inventory and VODA.ai’s tools help that effort. In this challenging time, we are excited to use the cutting-edge predictive modeling from VODA.ai to help us make more informed decisions, saving water and costs,” said Tim Thomure, Director of the Water Department. “There is limited data we currently have to make expensive decisions around maintaining our pipe network. With VODA.ai’s DaVinci machine learning technology, we will be able to make smart decisions save valuable resources, and protect our water infrastructure.”
Zella West, long-time Manager for the Nob Hill Water Association, said “Every utility has more miles of mainline that should be replaced than there is money in the budget. Nob Hill is using this program to direct our valve exercising program to the mains that are predicted to fail so that if they do fail, the damages can be kept to a minimum.” She added: “VODA.ai’s artificial intelligence platform finds patterns of pipe strengths and weaknesses for all of our water mains. They even predict which pipes are likely to fail within the next twelve months. This helps us make smarter decisions on pipes to replace or leave alone. Asset management decisions based on the age of pipes or their failure history are generally less than half as accurate as VODA.ai’s machine learning assessments.”
What are some of the next steps needed to advance the technology?
The biggest challenge for machine learning is the quality and quantity of asset data. VODA.ai has developed multiple algorithms to find and highlight missing data and anomalies. These are shared with utilities to verify or complete data sets. This results in improved data quality that benefits other digital systems used by the utilities.
LIFT can help the industry by validating the benefits of using machine learning to help utilities make smarter, proactive decisions for managing their buried assets. LIFT can also assist in finding partners to expand the application of machine learning to asset management. Finally, LIFT may be able to provide in-kind support for expanding initiatives for improving data quality and applying machine learning to smart decision making.Jim FitchettChief Operating OfficerVODA.ai