ProofBlue
EN JP
Request briefing
ProofBlue Industrial Intelligence · Read-only OT · Grounded AI · Private cloud

Industrial maintenance intelligence, grounded in plant data and OEM knowledge.

ProofBlue connects approved OT data through a read-only edge layer, runs governed maintenance workflows inside a private cloud boundary, and gives teams grounded AI answers backed by telemetry, OEM manuals, work history, and verifiable evidence.

Request a pilot briefing Access the guided demo
  • Connect PLC, SCADA, OPC UA, Modbus, MQTT, and API data without creating a control path back into the plant
  • Run the full Detect → Decide → Dispatch → Execute → Verify → Prove maintenance cycle
  • Ask asset-scoped AI questions that cross telemetry with OEM manuals, prior work, RCA, and verification outcomes
  • Protect industrial knowledge with encrypted private storage, tenant-scoped retrieval, and a private Bedrock path when enabled
Mikkeli membrane treatment pilot environment
Mikkeli reference site Real plant context. Real instrumentation. Real operating signals.

Built from an actual membrane-treatment environment rather than a presentation-only concept.

Digital MBR Mikkeli overview
One governed operating loop: Connect safely · Understand context · Diagnose with evidence · Execute maintenance · Verify the outcome · Preserve proof

ProofBlue Maintenance

A complete maintenance workflow — not another alarm dashboard.

ProofBlue keeps the operating context attached to every maintenance decision. Telemetry, alarms, diagnosis, work instructions, human approvals, execution evidence, and post-maintenance verification move through one governed lifecycle instead of becoming disconnected records.

01DetectIdentify an abnormal condition from alarms, rules, trends, or external events.
02DecideAcknowledge, diagnose, dismiss a false positive, or turn the event into governed work.
03DispatchDefine owner, approver, SLA, timing, checklist, parts, and verification window.
04ExecuteCapture tasks, readings, findings, notes, labor, materials, and attachments.
05VerifyCompare before and after under a defined window and decide whether the intervention worked.
06ProvePreserve the full timeline, source records, approvals, and Evidence Pack provenance.
Condition-based maintenance Human approval boundaries Typed work-order execution RCA + maintenance history CMMS / EAM integration-ready
Maintenance case · PB-MNT-000184 ● Grounded context
M-04
MBR Blower 04Mikkeli · Aeration train B
71Health
Detect · High priorityVibration trend above operating baseline
12 min
Vibration RMS8.4 mm/sLimit 7.1
Bearing temp.78 °C+11 °C / 24h
Operating load86%Stable
AI-assisted diagnostic draft Inspect drive-end bearing and verify lubrication / alignment before continued high-load operation.

Grounded in the active telemetry window, asset history, approved knowledge, and available OEM documentation. Human review remains required before governed action.

08:42 Condition detected and evidence window frozen
08:47 Operator acknowledged · diagnostic review opened
08:54 Draft Work Order prepared for approval

Grounded Maintenance AI

Ask what is happening to this asset — and require the answer to show its evidence.

ProofBlue Asset AI combines the exact asset context with authorized OEM manuals, telemetry analysis, active knowledge, Work Orders, RCA, and verification history. It is intentionally read-only: AI can explain, compare, summarize, and propose, but it does not silently change alarms, policies, PLCs, or maintenance records.

“Is this value outside the OEM operating range?”Live telemetry × model-specific limits.
“Why did this alarm happen now?”Event timing × operating state × documented behavior.
“Did the last intervention actually work?”Current condition × prior Work Order × verification outcome.
“What should the technician inspect next?”Asset state × OEM procedure × maintenance history.
✓ Cites sources ✓ Separates fact from inference ✓ Asset / tenant scoped ✓ Fails closed when evidence is insufficient
Ask ProofBlue · Asset scopedRead-only AI
TelemetryOEM manualWork OrdersRCA / history
Why is the bearing temperature rising while load is stable?
OEM FACT

The cited maintenance source defines the applicable inspection / operating guidance for this asset.

PROOFBLUE OBSERVATION

Load is stable while bearing temperature has increased over the active comparison window.

AI INTERPRETATION

The pattern is consistent with a bearing-side issue worth inspecting, but the available evidence does not establish root cause.

S1 · Telemetry analysisS2 · OEM sourceS3 · Prior Work Order

Private cloud by design

Industrial knowledge stays inside an explicit security and authorization boundary.

Designed for ProofBlue-managed AWS deployments and customer-controlled cloud profiles.

01Read-only plant edgeApproved OT signals only. No AI-created control path back into equipment.
→
02Private application VPCApplication services, authorization, retrieval, and maintenance workflows in private subnets.
→
03Encrypted data layerPrivate S3 + KMS for documents and PostgreSQL / pgvector for scoped knowledge retrieval.
→
04Private AI pathAmazon Bedrock through approved private networking when enabled; only authorized retrieved context is sent.
Tenant isolationTenant, site, asset, and model filters are enforced server-side before vector ranking.
Source governanceDocuments are versioned, classified, hashed, and linked to page / section provenance.
Human authorityAI can recommend and explain. Authorized humans approve maintenance policy and governed actions.
Evidence continuityHigh-value document, approval, execution, and verification events can be preserved in Evidence Vault.
Enterprise deployment path Private VPC · private data services · KMS encryption · least-privilege IAM · private Bedrock endpoint · optional customer-controlled AWS account / KMS boundary

ProofBlue Evidence Vault

Every industrial claim should lead back to evidence that can be checked.

Preserve calibration certificates, laboratory reports, commissioning records, photos, CSV files, and telemetry windows as immutable evidence records — linked to the exact sensor, asset, site, or event they support.

Standalone or telemetry-linkedStart with a document today. Connect it to live plant context when available.
Integrity that survives sharingSHA-256 fingerprints and ledger anchors expose silent file replacement.
Device passportsShow calibration validity, provenance, lifecycle, and related evidence through one record.
Explore Evidence Vault in the demo Discuss a sensor evidence pilot
Device passport● Ledger verified
◈
NH4 SensorHORIBA · SN H123456
✓ Evidence integrity verifiedCalibration certificate · valid until 11 Sep 2027
Evidence ID
EV-92A7D
Status
ACTIVE
Object
SHA-256 match
Anchor
Iroha found
Original→Hash→Record→Verify

Verifies provenance, integrity, status, and anchoring — not the intrinsic truth of every claim inside a document.

The problem

Plants generate data, but customers still struggle to use it as defensible evidence.

SCADA and historian systems are useful for operations, but they do not usually produce clean, shareable, customer-ready evidence for warranty reviews, compliance support, monthly reporting, or plant-performance communication.

Fragmented reporting

Operational data is spread across dashboards, registers, spreadsheets, and manually assembled reports.

Low external readability

Even good plant data is hard for municipalities, partners, and service stakeholders to interpret quickly.

Weak proof continuity

When issues arise, teams often reconstruct the past instead of using a structured, reviewable evidence trail.

The platform

ProofBlue adds an operational trust layer above the existing plant stack.

We turn raw plant telemetry into structured operational context, governed maintenance workflows, grounded AI, and proof-backed reporting.

Connect
Contextualize
Detect
Act
Prove
Wastewater performance assurance Monthly audit-ready evidence Warranty and compliance support Structured stakeholder reporting Built for membrane-treatment environments Grounded Maintenance AI Private-cloud deployment model

How ProofBlue works

One operating loop from signal to action and proof.

1
Connect

Approved PLC, SCADA, OPC UA, and Modbus signals flow through a read-only edge layer.

2
Contextualize

Signals receive units, quality rules, asset relationships, and operating context.

3
Detect

Configurable rules and trends identify abnormal behavior, alarms, and maintenance risk.

4
Act

Teams acknowledge events, prioritize interventions, and connect actions to CMMS or EAM workflows.

5
Prove

Telemetry windows, decisions, work performed, and documents become traceable evidence records.

Who it is for

Built for customer use cases first.

Operators and integrators

Need clearer monthly reporting, stronger differentiation, and less ambiguity around site performance.

Municipal utilities

Need transparent reporting, easier communication with stakeholders, and better visibility into treatment outcomes.

OEM and membrane-service teams

Need more trustworthy field context for warranty, support, and installed-base performance reviews.

Industrial water teams

Need structured evidence to support operations, reporting discipline, and process-performance communication.

Mikkeli solution model

Real wastewater engineering, configured for a concrete deployment scenario.

The Mikkeli solution model applies ProofBlue to a concrete membrane-treatment architecture, using mapped operating signals, edge connectivity, and defined reporting workflows.

  • UR32 / PLC / Modbus integration model
  • KPI streams including permeate flow, air flow, pressure, and MLSS
  • Backend ingest, MQTT bridge, persistence, and dashboard components prepared
  • Structured monthly proof-pack logic already defined
Instrumentation is real

The available configuration already maps concrete operating variables from the plant edge.

The data stack exists

Telemetry ingestion, storage, dashboards, and workflow components are already in place.

The value is practical

The customer outcome is clearer reporting and lower ambiguity — not extra system complexity.

Pilot paths

Start with one measurable operational outcome.

90-day Monitoring & Maintenance Pilot

For one site, line, or critical asset group.

  • Read-only telemetry integration
  • Signal and asset-context mapping
  • Configurable alarms and event workflow
  • Condition-based maintenance recommendations
  • Asset-scoped AI grounded in approved OEM / site knowledge
  • Baseline and outcome review

Evidence Vault Pilot

For sensor, calibration, warranty, or compliance evidence.

  • Device passports and evidence records
  • Document and telemetry linkage
  • Integrity verification and ledger anchoring
  • Public or permission-controlled sharing
  • Verifier-ready history

Full Site Assurance

Connect, View, and Audit working as one operational layer.

  • Multi-asset monitoring and reporting
  • Maintenance and alarm workflows
  • CMMS / EAM integration path
  • Proof Packs and stakeholder outputs
  • Ongoing operating support
Define a pilot outcome Explore the guided demo

Founders

Built by operators with field, platform, and Japan execution experience.

Luiz Costa

Luiz Costa

Co-Founder

Software engineer focused on turning complex operational data into reliable products and reporting systems. His work includes carbon-credit measurement projects deployed across multiple Brazilian sites and monitoring infrastructure for 642 energy-generation devices during Rio 2016.

LinkedIn profile
Timo Halonen

Timo Halonen

Co-Founder · Japan Execution

Tokyo-based operator with 30+ years in industrial IoT, utilities, and infrastructure. He brings deep Japan-market execution experience across wastewater, connectivity, channel partnerships, and customer development.

LinkedIn profile

Why ProofBlue

SCADA helps teams operate. ProofBlue helps them present performance with confidence.

What existing systems already do well

  • Operations visibility
  • Alarms and control context
  • Basic event history
  • Daily decision support

What ProofBlue adds

  • Customer-ready evidence packaging
  • Condition-based maintenance guidance
  • Alarm-to-action traceability
  • Performance reporting discipline
  • Warranty and compliance support
  • Cleaner stakeholder communication

Contact

Request a pilot briefing.

Use the form below to contact the ProofBlue team directly. We will receive the request by email and follow up with you.

This form sends your contact request directly to the ProofBlue team by email. See our privacy policy.

When submitted, the request is sent to ProofBlue and the download starts. See our privacy policy.