Presenting Baseform AI Assistant

Which 1,000 meters should I replace first, and what does it return in 12 months?

Ranked by **measured error × volume**, not by age: about **€78.6k a year**. Replacing the **oldest** 1,000 instead returns only **€6.7k**. Age is the usual rule, and it leaves over 90% of the value on the table.

Ranked by **measured error × volume**, not by age: about **$85k a year**. Replacing the **oldest** 1,000 instead returns only **$7.2k**. Age is the usual rule, and it leaves over 90% of the value on the table.

Which leaks are my alarms missing, and where do the crews go tomorrow?

Two zones have a minimum night flow about **20× their target** and **no event open**. Ranked, they are tomorrow's crew targets. Nobody is looking at them yet.

Which zones have the most pipe breaks, and where is it worst per kilometer?

By raw count, Zone 1 leads. Normalized by pipe length, the picture changes: {{Zone ; Pipe breaks ; Main (km) ; Breaks / 100 km // Zone 1 ; 853 ; 210.2 ; 406 // Zone 2 ; 677 ; 85.4 ; 793 // Zone 3 ; 487 ; 36.3 ; 1,342 // Zone 4 ; 408 ; 92.0 ; 444}} Zone 3 has fewer breaks than Zone 1 but more than three times the rate. That is where renewal pays first.

Which zones have the most pipe breaks, and where is it worst per mile?

By raw count, Zone 1 leads. Normalized by pipe length, the picture changes: {{Zone ; Pipe breaks ; Main (miles) ; Breaks / 100 miles // Zone 1 ; 853 ; 130.6 ; 653 // Zone 2 ; 677 ; 53.1 ; 1,276 // Zone 3 ; 487 ; 22.6 ; 2,160 // Zone 4 ; 408 ; 57.2 ; 715}} Zone 3 has fewer breaks than Zone 1 but more than three times the rate. That is where renewal pays first.

I can renew 20 km of pipe this year. Where do I avoid the most breaks?

Choosing by **predicted break rate** avoids about **1,200 breaks** in 10 years. By recorded breaks only: **680**. By picking the worst asbestos cement: **600**. The same 20 km, up to twice the return.

I can renew 12 miles of pipe this year. Where do I avoid the most breaks?

Choosing by **predicted break rate** avoids about **1,200 breaks** in 10 years. By recorded breaks only: **680**. By picking the worst asbestos cement: **600**. The same 12 miles, up to twice the return.

Why was this event detected?

The daily minimum flow crossed the detector's threshold (RQU ≥ 2.5) on **4 of the last 30 days**. These are the detector's current settings, which may have changed since. Explainable alarms mean faster triage.