HomeSolutionsServicesCommercialisationAboutInsightsTalk to SynapseAI

Industrial AI · Ireland · EMEA & Asia

Industrial automation, software and technology commercialisation.

For the people who run industrial plants, and for the companies selling technology into them.

Industrial AI · plant data to decisions Running

Industrial AI data flow: process signals from the historian, operations records from shift logs and work orders, and engineering knowledge from P&IDs, procedures and incident reports all feed a layer of models, retrieval and rules, which produces an operating decision with its evidence attached.

AI is the capability. The plant is the domain.
Software
Control-room tools. PID Tuner in service.
Consulting
Automation, industrial AI, digitalisation, OT security.
Commercialisation
Market entry and industrial sales, EMEA and Asia.
What we do

Three areas. One starting point: the operational problem.

Industrial software

Control-room tools built by control engineers.

See the software

Industrial consulting

Independent advice on where technology earns its place.

See consulting

Technology commercialisation

Market access for companies with good technology and no local presence.

See commercialisation
Product

SynapseAI PID Tuner

A recorded step test in. Tuning you can defend out — simulated before it reaches the controller.

01
Step test
02
Data quality
03
FOPDT model
04
Simulation
05
Tuning sets
06
Faceplate values
TIC-204 · closed-loop simulation Identified from the step test
SETPOINT 0 s 90 s
K2.41°C/%
τ18.2s
θ6.0s
0.994fit
ConservativeMs 1.4
Kc0.55   Ti26 s
RecommendedMs 1.7
Kc0.95   Ti20 s
AggressiveMs 2.0
Kc1.45   Ti15 s
SIEMENSABBHONEYWELLYOKOGAWAEMERSON DELTAVISA / SERIES / PARALLEL
Loop monitor · unit 200 Trending
TIC-204Reactor outlet
180.0°C
AUTO
PIC-102Column overhead
4.20barg
AUTO
FIC-301Feed flow
62.0t/h
AUTO
PID VALVE PROCESS TT SP MEASUREMENT FEEDBACK
Illustrative · loops of the kind the Tuner is used on
Available now
  • Nine data-quality checks before tuning
  • Valve stiction screening
  • Integrating and level loops
  • One-click MOC report
In development

Plant Memory

Somebody at this plant has seen this fault before. The record of it exists. At four in the morning, nobody can find it.

One machine, one night, and the evidence behind what was decided
RAISED LOGGED BY SEEN BEFORE FIXED BY JUSTIFIES Compressor K-204 EQUIPMENT High vibration ALARM · 04:12 Operator note SHIFT LOG Previous incident INCIDENT REPORT Maintenance record CMMS WORK ORDER Decision & reasoning DECISION
EQUIPMENT

Compressor K-204

The asset the question is about. Everything below attaches to this tag, not to a folder somebody has to find.

Select a node

The knowledge is not missing. It is scattered across systems that were never built to talk to each other, and it thins out every time somebody experienced leaves.

Nothing references anything else

The historian holds the trend. The CMMS holds what was replaced. The shift log holds what the operator actually saw, in their own words. The document system holds the P&ID. Four systems, one machine, no link between them. And nobody is going to migrate decades of records into a fifth system to fix that.

A search returns documents. An engineer needs reasoning

Keywords give you a list to read through under time pressure. What is actually needed is the thinking behind the last decision: what was tried, what was ruled out, and why the call went the way it did. That was never written down as a document. It sits implied across several.

Experience leaves faster than it gets written down

Every process plant runs partly on people who know what a particular machine does when it misbehaves. When they retire or move on, that goes with them, unless it was captured against the equipment tag rather than in a conversation at handover.

And when an incident is reviewed months later, or an auditor asks why the plant did what it did, the answer has to be shown rather than remembered.

Plant Memory is in development. We are not putting a number on what it saves. When there is evidence for one, we will publish it.

In development

Digital O&M Logbook

One shift, recorded so the next one can use it.

01
Shift start

Crew, state, open items.

02
Operator note

In the operator’s words.

03
Equipment status

Tagged to the asset.

04
Issue logged

Raised once, visible after.

05
Action assigned

Owner and expectation.

06
Handover

Structured, not prose at 06:00.

07
Audit trail

Evidence that stands up.

Shift record · unit 200 Nights
  1. 22:04 Handover in Crew of four. Two items carried over from days.
  2. 03:12 Issue raisedK-204 Vibration climbing on the outboard bearing.
  3. 05:55 Handover out Three open items, each owned. Signed by both supervisors.
3 of 12 entriesAudit trail complete

The strip above is the process. This is what it leaves behind — searchable months later, tagged to the machine rather than to a shift. Illustrative · in development

Commercialisation

From a good product to a repeatable route to market.

For technology companies whose constraint is distribution, not engineering.

TECHNOLOGY MARKET SEGMENT CUSTOMER PARTNER REVENUE
IRELAND EUROPE MIDDLE EAST AFRICA ASIA
Ireland

Company registered in Ireland; the work is delivered from here.

Select a region
Sectors

Industries we know.

Process

  • Oil & gas
  • Chemicals
  • Petrochemical
  • Cement
  • Metals & mining
  • Pulp & paper

Energy & utilities

  • Power generation
  • Water & wastewater
  • Renewables
  • Utilities

Manufacturing

  • Discrete manufacturing
  • Pharmaceutical
  • Food & beverage

Critical infrastructure

  • Data centres
  • Other OT-heavy sites
Evidence

Time inside the industry, not a study of it.

20+ yrs
Industrial experience
  • Automation engineering
  • Control systems & commissioning
  • Plant operations
EMEA · Asia
Where the work has been
  • Industrial technology sales
  • Business development
  • Channel development
Vendor-neutral
How advice is given
  • No reseller margin on hardware
  • Advice unshaped by supply
  • Including when it is “do nothing”
Process-plant control room with the plant overview, process flow diagram and alarm summary on the monitoring wall

It starts where the plant is run, and ends where somebody can show it worked.

Running a plant, or building the technology that runs it?

Tell us the problem. We will say plainly whether SynapseAI is the right fit.

Common questions

Questions we get asked.

What does SynapseAI do?

Three connected areas. It develops industrial software for process plants, starting with the SynapseAI PID Tuner. It provides independent consulting on automation, industrial AI, digitalisation and OT cybersecurity. And it helps industrial technology companies reach customers across EMEA and Asia through market entry, business development and direct sales.

Where is SynapseAI based, and which regions does it work in?

Based in Ireland, working across Europe, the Middle East, Africa and Asia. Software is delivered anywhere; consulting and commercial work is delivered remotely and on site depending on the engagement.

Which industries does SynapseAI work with?

Primarily the process industries — oil and gas, chemicals, petrochemical, cement, metals and mining, pulp and paper, power, water, pharmaceutical, food and beverage, and general manufacturing — plus data centres, renewables and utilities.

Is SynapseAI tied to a particular automation vendor?

No. SynapseAI is vendor independent. Its software works alongside whichever DCS, PLC, SCADA or historian a site already runs, and its consulting is not tied to reselling any single vendor's hardware.