Sanesul and 4Fluid: Data intelligence reduces scanning time by 6x and detects 126 commercial losses.

Anyone who works in loss management knows the size of the challenge: balancing the need for network scanning with the limited capacity of field teams. In a city with more than 33,000 calls, However, conducting a purely conventional proactive search is time-consuming and has high operational costs.

Historically, the average production rate (based on a baseline of 7,010 collections over 3 months) indicated that inspecting 5,000 calls it would take between 4 to 6 months. For the loss analyst, this long response time means having to live with the volume of unbilled water (ANF) running for a longer period.

We needed to change the game: we needed to surgical guidance to the field.

Targeted scanning with 4Fluid 

Between January 27th and March 10th, we implemented the 4Fluid to optimize this routine. The goal was to digitize the scanning, identify business losses (fraud), and detect suspicious points of leakage not visible in the network. 

The impact on productivity was immediate. What previously took up to six months was accomplished in just one month.  A significant gain in field efficiency and agility in decision-making.

Operational Performance and Accuracy in Numbers

The results of the case study show an accelerated learning curve for the system:

• Case Study Universe: 5,000 monitored calls.
• Business Losses (Theft or Fraud): 126 irregularities identified. 
• Physical Losses (Invisible Leaks): 76 suspected leak points mapped and confirmed for preventative maintenance.
• Efficiency gain: Reducing execution time by up to 83%.

The key advantage of artificial intelligence applied to losses is the refinement of the models. The system learned quickly from the characteristics of our network:

• January: 46% of assertiveness in alerts.
• February: It went up to 60%.
• March: It reached 66% accuracy On the field.

In other words, this means that two out of every three service orders (SOs) generated were successful. 

The Impact for the Future

For those who manage the water balance of the concession, this case proves that combating modern losses requires... predictive intelligence.

Increasing accuracy to 66% generates fuel savings, optimizes work orders for inspection/geophone teams and, most importantly, speeds up the process. Revenue recovery and reduction of the Loss Rate.

Technology does not replace management; it enhances decision-making.