How to Plan and Deploy an Industrial IoT Project
The comprehensive guide that replaces the vendor sales deck. Real talk on pilot traps, integrations, and scaling.
Reality Check: Most IIoT projects fail not because the technology doesn't work, but because they were scoped wrong, integrated poorly, or deployed without organizational readiness. This guide addresses all three.
1The IIoT failure statistics — and what's behind them↓
The failure rate: McKinsey estimates 70-75% of IIoT projects fail to scale beyond a pilot. The common reasons: pilot designed for technology demonstration rather than value delivery, no clear ownership of the data and outcomes, IT/OT integration underestimated.
The pilot trap: a successful pilot (controlled environment, motivated team, vendor support) often doesn't translate to scale (multiple sites, varying conditions, ongoing operations team support).
How to avoid it: design pilots explicitly for scalability, involve operations and maintenance teams from day one, solve a real problem that has a real champion.
2Phase 0 — Strategic alignment and use case selection↓
Use case prioritization: score each candidate use case on business value ($), technical feasibility, data availability, operational readiness, and integration complexity.
The business case: every IIoT project needs a business case with quantified ROI, payback period, and risk assessment.
Stakeholder mapping: who are the champions (operations leadership), the skeptics (IT security), and the bystanders (operations team who will ultimately use the system). All three need a strategy.
3Phase 1 — Current state assessment↓
OT network assessment: document existing PLCs, SCADA, historians, and communication infrastructure. Most facilities have little documentation of their OT network — this is the first step.
Data availability audit: what data is already being collected? What data exists in PLCs but isn't being used? What data requires new sensors?
IT/OT integration assessment: does your corporate IT network connect to the OT network? How? What security controls exist?
4Phase 2 — Architecture design↓
The reference architecture: sensors → edge → connectivity → platform → application. Design each layer before selecting products.
Cloud platform selection criteria: does it support your existing systems? What is the connector library? What are the data egress costs?
Network segmentation: the Purdue model and its modern interpretation. How to segment OT and IT networks while enabling the data flows needed for IIoT.
5Budget planning — the full cost of an IIoT deployment↓
Hardware & Software: platform subscriptions, application licenses, sensors, gateways.
Integration: the most consistently underestimated cost. OT system integration (connecting to PLCs, historians) is always more complex than vendors suggest.
The TCO calculation: most IIoT projects quote hardware and software prominently and bury integration costs. Always model 3-year TCO, not just year-one cost. Small pilot (10-20 assets) $50K-$150K, mid-size deployment (50-100 assets) $200K-$600K.