Deployment Roadmap

From Reactive to Autonomous in 18–36 Months

A phased deployment strategy that builds capability, proves payback, and scales across the fleet

Strategy

Three-Horizon Overview

Build capability, scale across fleet, then optimize the portfolio

Horizon 1

Build Capability

0–12 months

Stand up the platform, prove first predictive use cases, and achieve payback by month 12–18. Foundation APIs, core modules, and pilot plants operational.

Horizon 2

Scale Across Fleet

12–24 months

Multi-plant rollout across all technology types. Fleet-wide benchmarking, Computerized Maintenance Management System (CMMS) integration, and automated reporting. Standardized deployment templates.

Horizon 3

Portfolio Optimization

24–36+ months

Autonomization of routine decisions. Cross-fleet optimization, financial portfolio management, and continuous Machine Learning (ML) improvement from fleet-wide feedback loops.

Maturity

Deployment Maturity Tiers

Each plant progresses through three capability tiers

Tier 1: Operational Analytics

3–6 months per plant

  • Data connectivity & historian integration
  • Explorer & Trend modules deployed
  • Basic alerting & threshold monitoring
  • Team onboarding & workflow adoption

Tier 2: Predictive Maintenance

6–12 months per plant

  • ML (Machine Learning) models for key failure modes
  • Insight module & pattern library
  • CMMS integration (Dynamics 365)
  • Financial impact quantification

Tier 3: Fleet Reliability Optimization

18–36 months

  • Cross-plant benchmarking & optimization
  • Automated reporting & board dashboards
  • Autonomous agent workflows with approvals
  • Continuous ML improvement from fleet data

2026 North Star

By end of 2026, the fleet runs a single internal "Industrial Data + AI Platform" where:

  • SWAPP is the default workbench across all technologies
  • AI embedded in every workflow – detection to automation
  • Closed-loop CMMS – detection to resolution with financial tracking
  • Data as a product – domain-owned, governed, discoverable
  • Self-serve Data Mesh – federated access without central bottlenecks
  • Federated governance – domain autonomy with fleet-wide standards
  • Scadanerve – open sensors-to-agents alternative path
Architecture

Portfolio Logic

How the platform layers build on each other

1

Data Access APIs (Foundation)

piwebapi (.NET): fast REST API for operational data – visualization, summaries, extraction

pifastapi (Python): analytics API for statistics, ML, batch processing

pipolars: high-performance Polars SDK + industrial transforms (used by pifastapi)

2

Product Platform (Work App + Extension system)

swapp (core): auth, navigation, workspaces, saved views, API Gateway (routes to piwebapi/pifastapi)

3

7 Application Modules

explorer, trend, stats – core reliability workflows

alert, insight, report – predictive intelligence, performance tracking & business reporting

4

Strategic Independence (PI alternative path)

scadanerve: sensors-to-agents pipeline (OPC UA/Modbus/MQTT – historian – feature store – agents)

Timeline

Quarter-by-Quarter Roadmap

Detailed deliverables and success metrics for 2026

Ready

Ready to Use – Foundation APIs & SDK

Available now

Status: Production-ready data access layer powering all SWAPP modules and analytics workflows.

piwebapi (.NET)

  • Fast REST API for operational data: visualization, summaries, extraction
  • Standard endpoints: points, summaries, recorded, interpolated, metadata, AF mapping
  • Observability: logs/metrics/traces + SLIs (latency, error rate)

pipolars (SDK)

  • read_pi() and read_af() wrappers returning Polars LazyFrames
  • Time zone handling, resampling, interpolation, quality flags, unit normalization
  • "Industrial transforms": rolling stats, event windows, downtime masks
Q1

Platform + Explorer

January – March 2026

Goal: Launch the platform core with first application modules. Establish Single Sign-On (SSO) and Role-Based Access Control (RBAC).

pifastapi (Python)

  • Analytics API: batch queries, statistical summaries, resampling
  • Integration with pipolars for Polars-based transforms
  • Endpoints for aggregations, feature extraction, ML pipelines

SWAPP core

  • SSO (Single Sign-On) / auth, role-based access, workspace concept, saved queries, export
  • API Gateway routing to piwebapi/pifastapi based on use case
  • Extension framework (how modules plug in)

explorer.swapp

  • AF tree browsing, favorites, search, attribute/element compare
  • "Open in Trend" & "Open in Stats" deep links

AI baseline

  • "Data Copilot": natural language – PI query builder (safe, explainable)
  • Prompt + template repo (versioned), usage logging, red-teaming basics

Success Metrics

  • First SWAPP users can browse assets and pull timeseries without PI UI
  • <1% failed requests on normal load
  • Explorer adopted by pilot plant teams
Q2

Trend + Stats as Daily Driver

April – June 2026

Goal: Deliver the "daily driver" trending and analytics modules as the primary analysis interface.

trend.swapp

  • TrendMiner-like UX: multi-tag plots, time shift compare, overlays, annotations
  • Context layers: alarms/events/maintenance windows
  • Team sharing: "saved views", comments, snapshot links

stats.swapp

  • Guided analysis flows: correlation, regression, outliers, distributions
  • Capability & control charts where relevant
  • Notebook-like "analysis recipes" saved as reusable templates

AI inside Trend

  • Auto-insight: "what changed?" (change-point detection + narrative)
  • "Explain this trend": generates hypotheses + shows supporting evidence links

Success Metrics

  • 50-150 recurring internal users (operations/performance)
  • SWAPP Trend used in at least 3 recurring meetings (daily/weekly)
  • 20% less dependency on licensed trending tool for pilot plants
Q3

Alert & Insight

July – September 2026

Goal: Launch predictive modules and connect detection to business workflows.

alert.swapp

  • Anomaly detection & threshold alerting
  • Event correlation & alert routing
  • CMMS integration (Dynamics 365 dispatch)

insight.swapp

  • Failure mode classification & pattern library
  • Similarity search & RCA knowledge base
  • Fleet-wide pattern detection

AI inside Alert & Insight

  • "Ask the dataset" (Retrieval-Augmented Generation (RAG) over metadata + safe compute functions)
  • Auto-generated reports (weekly plant summary) with citations to charts/queries

Success Metrics

  • First predictive alerts generate CMMS work requests
  • At least one monthly report generated primarily from SWAPP
  • Pattern library seeded with fleet-wide failure signatures
Q4

Scale + report.swapp

October – December 2026

Goal: Scale the platform and launch business reporting across the fleet.

report.swapp

  • Time-based reporting (weekly/monthly) with loss accounting
  • KPI dashboards (availability, efficiency, heat rate)
  • Financial impact quantification per predictive action
  • Power BI integration for board dashboards

Platform hardening

  • Multi-plant scaling, tenant boundaries, cost controls, strong auditing
  • Data governance: catalog, lineage, classification, retention rules
  • MLOps (Machine Learning Operations) / LLMOps (Large Language Model Operations): model registry, evaluation harness, prompt/version governance

Success Metrics

  • SWAPP is default interface for at least one technology fleet
  • Financial impact of predictive actions tracked in Power BI
2027 Vision

Scadanerve – Open Sensors-to-Agents Pipeline

Strategic independence from proprietary historians with an open, agent-driven data pipeline

Edge Ingestion

OPC UA / Modbus / MQTT streaming from plant sensors with edge preprocessing and buffering

Open Historian

Open-source time-series store with AF-like asset model, replacing proprietary historian dependency

Agent Layer

Autonomous anomaly detection and alert explanation agents running per asset class

Bridge Mode

Scadanerve and PI running side-by-side, validating data parity before full transition

Feature Store

Real-time and batch feature engineering pipeline feeding ML models directly from the open stack

Fleet Mesh

Multi-plant Scadanerve federation with centralized governance and cross-plant data products

Scadanerve starts as a real plant pilot slice in 2027, proving the full sensors – agents – open time-series path

AI Strategy

AI-First Capabilities

What "AI-first" means in SWAPP – implemented in three tiers

Tier A

Assist

Fast wins
  • Natural Language (NL) – query builder (PI/AF search, time windows, summaries)
  • Auto chart descriptions + anomaly highlights
  • "Explain differences" between periods/plants/units
Tier B

Analyze

Repeatable intelligence
  • Root Cause Analysis (RCA) assistant: hypothesis generation + evidence scoring
  • Pattern library: recurring failure signatures (vibration, stator temp drift, etc.)
  • Automated feature extraction pipelines (pipolars recipes)
Tier C

Act

Agents with guardrails
  • CMMS dispatch: auto-create work requests from detected anomalies
  • Runbooks: "if X then do Y" with human approval gates
  • Continuous monitoring agents per asset class
Evolution

Digital Twin Maturity Model

SWAPP at DT3 (Predictive), moving toward DT4 (Optimization)

DT0 Standalone Manual data collection, no digital connection
DT1 Descriptive Real-time data visualization, historian-connected
DT2 Diagnostic Root cause analysis, pattern recognition
DT3 Predictive ML-driven failure prediction, closed-loop Computerized Maintenance Management System (CMMS) — SWAPP today
DT4 Optimization Fleet-wide autonomous optimization — SWAPP target
DT5 Autonomous Self-healing systems with full autonomy
Governance

Phase Gates

Key decision points tied to outcomes

End Q1 Gate

explorer.swapp adopted by pilot plants; platform core operational with SSO and API Gateway

End Q2 Gate

trend.swapp and stats.swapp adopted as daily drivers; 50+ recurring users; 20% less dependency on licensed tools

End Q3 Gate

alert.swapp and insight.swapp in use; first predictive alerts dispatched to CMMS; financial impact of avoided outage documented

End Q4 Gate

Multi-plant scaling proven; report.swapp live with Power BI; total avoided-cost tracked in dashboards

See the Platform in Action

Explore use cases and ROI projections for each technology