ChileChemicals and materials2026 – ongoing

Sociedad Química y Minera de Chile (SQM)

Predictive models for nitrate-production operations

AVEVA reported that SQM cut manual analysis effort by 30% to 50% in nitrate operations

WorkingVendor case study

Problem

SQM needed better access to operating data to improve nitrate production. Gathering information and analyzing plant conditions delayed operational decisions.

Approach

SQM connected plant operations to cloud analysis using AVEVA PI System, CONNECT and PI Vision. Predictive models generated recommendations that returned to the plant for operators to use. AVEVA reported that information access fell from days to minutes.

Results

30%–50%reduction in manual analysis effort reported by AVEVAVendor case study[1]

around 90%predictive-model accuracy reported by AVEVA, without a stated evaluation methodVendor case study[1]

Takeaway

The workflow returned model recommendations to plant operators, linking prediction with operating decisions.

Sources

SQM optimiza producción de nitratos y logra precisión de modelos predictivos en torno a 90% con soluciones de AVEVA
G5noticias20 Aug 2026News article

Source summaries

[1]G5noticias20 Aug 2026

G5noticias reported AVEVA's account of predictive modeling in SQM's nitrate operations. It gave model accuracy of around 90% and a 30% to 50% reduction in manual analysis effort. It described a workflow that returned recommendations from cloud analysis to plant operations.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026