New: Networks — CCTV surveys, condition grading and renovation programmes for water and sewer networks, in the same service. Explore Networks →

Features

Every Pipelytica feature across the two products: what is in use and what is coming, and who decides in each.

What the statuses mean

  • In use used by customers today.
  • In pilot built, enabled for pilot customers after a demo.
  • Coming not available yet.

Shared foundation

The same video analysis, the same way of reviewing and the same security in both products.

13 in use

Analysis and review

AI analysis of CCTV videoIn use
The AI goes through the inspection video and suggests observations with screenshots. A person reviews every suggestion before the report.
Who decides: the AI suggests, a person confirms
Metre reading from the videoIn use
The distance counter burned into the video is read with text recognition, so every observation gets its position in metres.
Who decides: runs on its own
EN 13508-2 based codesIn use
Observations are recorded with codes based on the EN 13508-2 standard, and a person confirms the code.
Who decides: the AI suggests, a person confirms
Severity grade 1–4In use
The AI suggests a severity grade of 1–4 for every observation, and a person confirms it.
Who decides: the AI suggests, a person confirms
Capture-condition detectionIn use
Water, fog, reflections, LED light artefacts and dark frames are recognised so they do not become defect observations. This does not remove every false alarm.
Who decides: runs on its own
Analysis queueIn use
Videos are analysed in a queue that takes turns between companies, and an interrupted analysis returns to the queue on its own.
Who decides: runs on its own
Model training only with permissionIn use
Approvals, rejections and corrections are stored. They are used to train the AI only if the user has given permission in the settings. By default there is no permission.
Who decides: a person

Users and security

Four user rolesIn use
Administrator, office, inspector and customer viewer, each with its own rights. Customer viewers are free.
Who decides: a person
Company-level isolationIn use
Each company only sees its own data, and the server checks this on every request.
Who decides: runs on its own
Audit logIn use
Key actions are logged: who did what and when. The administrator sees the log in the settings.
Who decides: runs on its own
Secure sign-inIn use
Two-factor authentication, and sign-in is blocked for a while after repeated failed attempts.
Who decides: runs on its own
Finnish and EnglishIn use
The whole interface in Finnish and English.
Who decides: runs on its own
Billed by invoiceIn use
The service is billed by e-invoice or a PDF invoice by email. No credit card needed.
Who decides: runs on its own

How the AI is measured

We do not publish validation scores from our own training data. We measure the finished product on videos the model has never seen, against the labels of an inspector.

Measured results

September 2026 · 34 field videos · labelled by an EN 13508-2 inspector

MeasureResult
Right defect family in the customer report

The name of the finding (e.g. corrosion or sediment) matches the label the inspector gave. 95% confidence interval 51–69%.

60%
Exact defect type right

The same, with deposit sub-types (sand, scale, sludge) told apart. Inspectors often disagree on these too.

39%
Marked defect spots found

Share of the defect spots the inspector marked where the AI produced a finding. The rest are left for the inspector to find.

69%
Analysis time relative to video length

Measured on the production server: a 67-second video analysed in 45 seconds.

0.7 ×

How it was measured

34 field videos were kept entirely out of model training. An EN 13508-2 inspector marked the defect spots, and every AI finding was compared with those marks. The confidence interval is computed per video.

Why not 100%?

In published studies even experienced inspectors miss about a quarter of defects and code the same defect differently (Dirksen et al. 2013). The AI is a second pair of eyes, not a judge.

AI can make mistakes – Always verify findings before delivering a report to a customer.

Try it with your own video

Start a free 14-day trial. No credit card, no commitment.