Skip to content

Manufacturing · Vertex Manufacturing

Predictive maintenance saves manufacturer $3.1M in downtime

A precision-components manufacturer cut unplanned downtime 38% with IIoT telemetry and failure-prediction models across 3 plants.

38%

unplanned downtime reduction

$3.1M

annual downtime savings

27%

maintenance spares cost cut

14 days

average failure warning lead time

The Problem

Vertex's CNC and stamping lines suffered unplanned stoppages costing ~$40k per hour. Maintenance was calendar-based — over-servicing healthy machines while missing early failure signals on others.

The Solution

RippleCode deployed vibration, temperature and power sensors across 240 machines, streaming telemetry into an AWS IoT pipeline. Anomaly-detection and remaining-useful-life models trigger work orders in the existing CMMS, and plant dashboards give supervisors real-time asset health scores.

Architecture

  • Edge gateways with local buffering across 3 plants
  • AWS IoT Core + Kinesis streaming ingestion
  • Time-series feature pipeline and anomaly detection (Prophet + autoencoders)
  • Remaining-useful-life models per machine class
  • CMMS integration for automated work-order creation

ROI

Full program cost recovered in 7 months; ongoing savings of $3M+ annually across downtime, spares and overtime.

Engagement

Client
Vertex Manufacturing
Industry
Manufacturing
Service
AI & IoT
Timeline
10 weeks pilot (one line), 6 months to all 3 plants

Technologies

AWS IoT CoreKinesisPythonTimescaleDBGrafanaSageMaker

We went from firefighting breakdowns to planning maintenance two weeks ahead. The ROI case made itself after the first prevented line-down event.

Elena Rodrigues

VP Operations, Vertex Manufacturing

Facing a similar challenge?

We'll walk you through how this architecture would apply to your environment — free of charge.