Log in Request a demo

FAQ

Frequently asked questions

What the model needs, what it returns, and what it takes to get started.

About FractaLeak ZonesYour dataWhat you receiveWorking with usRisk assessmentAs-Built Digitization

What Fracta does

3 questions
Who is Fracta for, and what does it help us do?

Fracta serves water utilities and the engineering firms that support them. We help you find leaks faster with Fracta Leak Zones, prioritize which mains to inspect, maintain or replace with our risk assessment (Likelihood of Failure, Cost of Failure and Total Risk), and turn legacy drawings into GIS-ready data with As-Built Digitization.

What is the difference between Likelihood of Failure, Cost of Failure and Total Risk?

Likelihood of Failure (LOF) estimates a pipe’s probability of failure over a defined period. Cost of Failure (COF) estimates the financial impact, including repair and disruption, using specified cost assumptions. Total Risk combines the two to help compare priorities across the network.

Do we need new sensors or field inspections?

No. Our risk assessment uses your existing asset and break records plus environmental and geographic data Fracta maintains. Fracta Leak Zones adds an optional SCADA export, never new hardware or a live connection. Both tell your crews where to focus field work; they do not measure a pipe’s physical condition.

Fracta Leak Zones

9 questions
What is Fracta Leak Zones?

An AI-driven service that ranks the areas of your network most likely to contain leaks, so crews know where to start. It combines Fracta’s Likelihood of Failure model with your operating data.

How is this different from a traditional leak survey?

A survey sweeps miles of pipe with no clear starting point. Fracta Leak Zones tells you which areas to survey first, so crew hours go where leaks are most likely. Your survey team or partner still confirms each leak in the field.

What data do you need?

Only a GIS pipe inventory (pipe ID, install date, material, diameter) and break history are required. For the sharpest targeting, we recommend a SCADA export of hourly flows, tank levels and pressures, plus GIS of pressure zones and facilities.

Does Fracta Leak Zones need access to our SCADA system?

No. We only need a CSV or Excel export 1 to 4 times a year, matching your package. It is recommended, not required.

Our data has gaps. Can we still move forward?

Yes. Send what you have and flag what is missing. We will tell you up front how gaps affect scope.

Do we need to buy the risk assessment first?

No. Fracta Leak Zones is a standalone service and does not require a prior Fracta risk assessment.

How often are priority areas refreshed?

Once, twice, three or four times a year (Bronze, Silver, Gold or Platinum), so your team has time to investigate and act on each set before the next one arrives.

Who confirms the leaks?

Your crews confirm leaks in the field. If you prefer, Fracta can include on-site pinpointing through a leak detection partner as an optional add-on.

How is Fracta Leak Zones priced?

Pricing is based on the miles of main analyzed and how many deliveries you receive each year. Optional on-site pinpointing is quoted separately. Contact us for a quote.

Your data

3 questions
What data do we need for a risk assessment?

Start with your pipe-network GIS data, including unique pipe IDs, material, diameter and installation year, plus dated break or repair records. Breaks need a pipe ID or a usable location. Fracta reviews the files and identifies gaps before scoping the assessment. Fracta Leak Zones needs the same two datasets to start. A SCADA export and zone/facility GIS are recommended, not required.

Can we get started with incomplete records?

Missing pipe attributes do not automatically rule out an assessment. Some values can be estimated or recovered from supporting records. We also need to review whether the available break history and asset locations are usable. Estimated attributes should not be treated as field-verified facts.

Can Fracta convert drawings into GIS data?

Yes. Our As-Built Digitization service converts scanned drawings and legacy maps into georeferenced features and structured attributes for GIS use. We review the source records and agree on the required layers, attributes and output format. Digitization represents what the records show; it is not a field survey.

What you receive

4 questions
What will our team receive?

For a risk assessment, your team reviews pipe-level results on an interactive map, examines contributing factors, and uses reports and data exports. For Fracta Leak Zones, you receive a map of priority areas for leak investigation, refreshed 1 to 4 times a year. For As-Built Digitization, you receive GIS data in the agreed structure and format. The exact analyses depend on the agreed scope.

Can we use the results in our existing GIS?

Yes. Assessment data can be exported in formats such as CSV, GeoJSON and Shapefile for use in existing GIS and analysis workflows. File export is different from a live integration; specific connections or automated data exchange should be confirmed during scoping.

How can we evaluate the predictions?

Fracta evaluates predictive performance using historical break records held out of model training. Ask to review the validation approach and results relevant to your assessment. Outputs express probabilities, not guaranteed failure dates, and should be used alongside engineering judgment and field evidence.

Can Fracta help with capital planning?

Yes. Risk results help compare replacement priorities. Long-term projections can explore how different renewal rates affect expected breaks. These outputs support planning decisions; they are not construction designs, and projections depend on the data and assumptions used.

Working with Fracta

3 questions
What does working with Fracta involve?

We start by understanding your goals and reviewing your available data or drawings. We then agree on the scope, prepare and check the data, carry out the analysis, leak prioritization or digitization, and review the deliverables with your team. Your local knowledge helps resolve gaps and interpret the results.

How long does a project take, and how is it priced?

Timing and pricing depend on the work requested and the condition and volume of the source data. For risk assessments, network size helps define scope; for digitization, the record set does. We review those inputs before confirming a quote and delivery schedule. For Fracta Leak Zones, scope is set by network size and how many deliveries you receive each year.

What happens after delivery?

Your team can use the in-app user guide for operating instructions and contact Fracta for assistance. For risk assessments, new break records, asset changes and model updates can affect results. Confirm the support arrangements, platform access and reassessment schedule included in your agreement. With Fracta Leak Zones, updated priority areas arrive on your package schedule, so your team can act on each set before the next one lands.

Risk assessment: Likelihood, Cost and Total Risk

39 questions
What does Fracta do?

Fracta scores every pipe segment in your distribution network for how likely it is to fail and what that failure will cost, so your next capital dollar replaces the segment carrying the most risk. It uses machine learning on the records you already have. No new sensors, no field work.

How do the three scores fit together?
ScoreQuestion it answersUnit
Likelihood of Failure (LOF)Which pipes are most likely to break?Probability (%) at 1, 2, 3 and 5 years
Cost of Failure (COF)What happens, in dollars, when they do?$ per segment (direct + indirect)
Total Risk (TR)Where should the next capital dollar go?$ at risk per segment (LOF × COF)
Which networks can you score?

Primarily potable water mains. Reclaimed water, transmission lines and others can be included by agreement.

How is this different from ranking by age or break history?

Age, material and break history are three of the many variables that decide whether a pipe fails. On their own they send crews to the oldest pipe, not the weakest one, and they say nothing about what a break costs. Fracta weights 120+ variables learned from 560,000+ recorded breaks, instead of weights assigned by hand.

What results have utilities seen?

Topeka, KS (three engineers, $6.5M in annual waterline replacement) saw 54% of pipe failures fall in the highest-risk tier in year one and 65% in year two, with an average $18K direct savings per capital project and $80K to $100K in annual savings. Fracta is trusted by over 100 utilities across North America and has analyzed 200,000+ miles of pipe.

Topeka, KSResult
Failures caught in the highest-risk tier, year one54%
Failures caught in the highest-risk tier, year two65%
Average direct savings per capital project$18K
Annual savings$80K to $100K
What is Likelihood of Failure?

A machine learning score of each pipe segment’s probability of failure at 1, 2, 3 and 5 years, plus its percentile rank across your system and the five variables contributing most to that score.

What goes into the model?
  • Pipe inventory from your GIS: material, diameter, install year, length
  • Break and repair history on the segment and its neighbors
  • Soil chemistry and corrosivity
  • Frost depth, elevation and surface conditions
  • Land use, population served and operating context
Is the model trained on our data or an industry average?

Both. It is trained on 560,000+ recorded breaks across 200,000+ miles of pipe, then calibrated against your own system’s failure record.

What are the Top 5 Risk Drivers?

For each segment, the five factors contributing most toward a higher predicted LOF. The order is learned from data, not set by a fixed rule like “age always comes first,” so it differs pipe to pipe.

If soil ranks above age, does that mean the pipe is corroding?

No. It means soil-related inputs contribute more to that pipe’s predicted LOF than age does, given the other inputs. A driver explains the prediction; it does not prove the physical cause or measure present condition. Drivers are relative influence, not percentages of failure probability.

How accurate is it, and how can we check?

Fracta evaluates predictive performance on historical break records held out of model training. At Topeka, 54% of failures landed in the top-risk tier in year one and 65% in year two. Outputs are probabilities, not guaranteed failure dates.

Does the model improve over time?

Yes. Accuracy improves as your break data accumulates. With a Fracta Monitor subscription, your network is re-scored annually as your data changes.

What is Cost of Failure?

The dollar cost of a break on each segment, split into direct cost (repair and restoration) and indirect cost (the consequences of the break). It is location aware: a break under a quiet side street and one on an arterial outside a hospital are not the same event.

What costs are included?
Direct costIndirect and consequential cost
Excavation, repair and pipe replacementService interruption to downstream customers
Pavement and surface restorationTraffic disruption and detour duration
Emergency crew mobilization and overtimeImpact on hospitals, schools, emergency services
Intense repair (HDD, jack and bore) near rail and major routesWater lost, environmental exposure, property and third-party damage
Where do the default cost values come from?

Direct costs come from AWWA handbooks, Water Research Foundation findings, public bid results from small, medium and large cities, and academic cost studies. Indirect defaults use public bid documents for traffic control, environmental fine records and WRF technical reports. Defaults are refined over time with anonymized values entered by other Fracta customers.

Can we use our own costs?

Yes, and we recommend it. Every input can be edited, and a custom COF field lets you add your own rule by diameter, material or install year range. Property damage and critical facility costs have no defaults, because public data is too thin, so your team supplies those.

How do you estimate customers affected?

Three methods, from simplest to most precise:

  1. Standard: census population density within a 50 m buffer of the pipe.
  2. Valve model: uses your valve layer to find which adjacent pipes are isolated when a pipe fails.
  3. Hydraulic model: you upload results (customers out of service and below 20 psi per pipe) from your own hydraulic model as a CSV.
How are critical facilities and environmental risk counted?

A geospatial buffer around each pipe counts hospitals, schools, fire and police stations, streams and water bodies from Fracta’s in-house database. Each counted feature adds its input cost. You can view these features in the COF map view.

What are COF scenarios?

Named sets of cost assumptions you create in the platform (Scenarios tab) and switch between, so you can compare, for example, a conservative and an aggressive cost view. COF and Total Risk update with the selected scenario.

What is Total Risk?

Likelihood of Failure multiplied by Cost of Failure: the expected dollars at risk on each segment. It is one objective criticality score, comparable across the whole system and readable by anyone who approves a budget.

Why not just replace the pipes most likely to break?

Because a likely break can be cheap and an unlikely one can be expensive. In this example, condition alone sends crews to Segment A first; Total Risk moves Segment B 35 places earlier.

Segment A (residential side street)Segment B (arterial near hospital)
LOF, 5-year82%31%
Cost of Failure$9,000$75,000
Total Risk$7,380$23,250
CIP rank#41#6
What is Total Risk used for?

Building a CIP ranked in dollars, sizing budgets by how much risk each funding level buys down, and supporting master plans, rate cases and grant applications with an auditable figure per segment.

What do we receive from a risk assessment?

Total Risk in dollars for every segment, a system-wide ranking, a map layer symbolized by Total Risk, an exportable ranked asset list for the CIP, COF scenario comparison, and Job Planner grouping for project bundling.

Can you project how breaks change under different renewal budgets?

Long-term projections can explore how different renewal rates affect expected breaks. They support planning decisions; they are not construction designs.

What data do you need for a risk assessment?
DatasetRequired?Key fields
Pipe assetsYesUnique ID, install date or year, material (with code legend), diameter; status optional
Break historyYesBreak date on every record; pipe ID, geometry, lat/lon or address
Supplementary layersOptionalPressure, parcels (year built), valves, hydrants, reservoirs/plants, joints

Most utilities already have 90% of this in their GIS.

How do we know if we’re ready?

Take the six-question Data Readiness Check (about 15 minutes). Most utilities clear five of six from their GIS alone.

Can we start with incomplete records?

Yes. Blank values are workable; missing columns are not. Gaps in install year, material or diameter can be recovered from connecting pipes and supplementary layers. Estimated attributes should not be treated as field-verified facts.

What happens to breaks without dates or locations?

Undated breaks are removed from the model. Breaks without a pipe ID are matched to the nearest pipe. Addresses are geocoded; incomplete or erroneous addresses may be dropped.

What formats, units and projections do you accept?

Shapefile, GeoPackage, File Geodatabase or GeoJSON; breaks can also come as Excel or CSV with locations. Imperial or metric, any projection or national grid, local-language field names. Tell us the units and CRS; no need to rename, translate or clean anything.

How do we send data, and what happens next?

Export the layers, send them via the secure link your Fracta contact provides or your utility’s SFTP, and Fracta returns a written completeness summary flagging missing fields and anomalies before scoping.

Our pipes aren’t in GIS yet. Can we still do this?

Yes, through As-Built Digitization, which converts record drawings into GIS data first. See the As-Built Digitization questions below.

What will our team see in the platform?

A web map of your network with Assets and Risk Analysis views: LOF, COF, Total Risk, length and population layers (depending on subscription), plus Population and Critical Infrastructure context layers. Filter by any property or a drawn area, click any pipe for details, and manage COF scenarios.

How is the risk assessment priced?

Per mile of main, banded by system size: larger systems pay less per mile. Every quote is built from a verified mile count. The package is a one-time baseline assessment of the whole network with year one of platform access included.

What happens after year one?

From month 13, a Fracta Monitor subscription keeps platform access and re-scores your network annually as your data changes. A three-year term is recommended; one- and five-year terms are available.

Can we buy through a cooperative contract?

Yes, cooperative purchasing vehicles are supported where a utility buys that way.

What does the process look like?
  1. Readiness check: about 15 minutes.
  2. Send your GIS export and break history.
  3. Data review and scope: from a verified mile count.
  4. Model build and QC: calibrated to your failure record.
  5. Live results walkthrough: with your engineers, then the ranked CIP export.
How long does it take?

Timing is confirmed after we review your data. Historically, assessments have taken about four to eight weeks; data quality and follow-up requests can affect the schedule.

How is our data protected?

Data can be transferred through the secure link provided by your Fracta contact or your utility’s SFTP. Before transfer, confirm the hosting and access arrangements, retention and deletion terms, and any utility-specific security requirements with Fracta.

What support do we get?

An in-app user manual, your Fracta representative and contact@fracta.ai.

As-Built Digitization

18 questions
What is As-Built Digitization?

It converts as-designed and as-built drawings (scanned record sets, mylars, vector PDFs and CAD files) into GIS data that is georeferenced to a coordinate system, connected at junctions and attributed to your schema. The service is patent pending.

We already have a GIS. Why would we need this?

Most utilities do, but the authoritative record of what is actually in the ground often still lives in drawings, not the GIS. Digitization closes that gap so your GIS reflects your records, not just what someone had time to draw in.

Which networks do you support?

Water distribution, wastewater collection, stormwater, gas distribution and transmission, liquids pipelines, electric distribution, and telecom and fiber outside plant.

Who typically owns this work?

GIS managers and analysts, records and asset managers, distribution or collection system managers, engineering PMs at capital project closeout, city engineers and county surveyors, gas integrity leads, outside plant engineering teams, and GIS leads at engineering firms.

What is driving utilities to do this now?
SectorRequirementDate
Drinking waterLead and Copper Rule Improvements service line inventoriesNov 1, 2027
Gas transmissionPHMSA TVC records and MAOP reconfirmation (49 CFR 192.624)Jul 3, 2028
Wastewater / stormMS4 outfall mapping; consent decree deliverablesPer permit
Electric / public worksEsri Utility Network migration after ArcMap retirementOngoing
Telecom / broadbandFCC Broadband Data Collection; BEAD as-built reportingPer program
Engineering / SUEASCE 38-22 records researchPer project

Dates reflect federal rules as published; confirm state-level variation with your regulator.

Do we need to buy the risk assessment to get digitization?

No. As-Built Digitization is sold and priced entirely on its own. It is also the natural first step if you later want Likelihood and Cost of Failure scoring, which requires the network in GIS.

What drawing formats do you accept?
  • Scanned raster drawings (TIFF, PDF, JPEG)
  • Vector PDF and GeoPDF
  • CAD files (DWG, DGN)
  • Mylar and paper sheet scans
  • As-designed and as-built record sets
  • Mixed-vintage sets, including unindexed archives
We only have paper or mylar. Do we have to scan them first?

Yes. Our workflow starts with digital files, so paper and mylar need to be scanned first. Contact Fracta before arranging scanning to confirm file requirements for your drawings.

How do we send you the drawings?

We use file-based exchange through a sharing method agreed with your team, including your organization’s existing file-sharing service. Contact Fracta to arrange the transfer. No connection to your live GIS is required.

How does it work?
  1. Centerline vectorization: raster and vector drawings become continuous centerline networks, connected at junctions and ready for GIS analysis, hydraulic modeling and asset management.
  2. Automated georeferencing: OCR and language models read street labels, grid ticks and sheet annotation. Affine and thin plate spline transforms align each sheet to State Plane or UTM.
  3. Attribute extraction: on-sheet text and callouts are parsed into structured attributes (diameter, material, install year, status) and standardized to your existing schema.
Is this fully automated, or do people check it?

There is one deliverable quality level: human-verified and GIS-ready.

How accurate is the output?

Every delivery includes a QA report with residual error at control points, so positional accuracy is measured, not asserted.

Is digitized data the same as a field survey?

No. Digitization represents what the records show. If the original drawing is wrong, the GIS will faithfully reflect that error. It is a strong basis for records research, not a substitute for field verification.

Can we start with a pilot?

Yes. A small sample can be digitized at no charge so you can review the output before a larger engagement. Agree on the sample size and scope with Fracta first.

What do we receive?
  • Esri File Geodatabase
  • Utility Network ready feature classes
  • Esri asset package
  • GeoJSON and Shapefile
  • Attribute tables standardized to your schema
  • QA report with residual error at control points
  • Optional SVG overlay for visual review
What does a real delivery look like?

From one municipal storm drainage set (vector PDF), the delivered layers included:

LayerFeatures delivered
Storm drain structures52
Storm pipe segments49
Sanitary manhole covers17
Water valve boxes10
Control points carried for QA5
How is digitization priced?

Per sheet, banded by volume: more sheets, lower cost per sheet. Every quote is built from a verified sheet count, never a number guessed on a call. A one-time service, with no required Fracta subscription for the delivered GIS files.

Who owns the digitized data?

You retain rights to your source drawings and the GIS datasets derived from them.

Still have a question?

Ask us directly, or see the model run on your own network.