The Automation Company

Process

Audit. Retrofit. Integrate. Operate.

Every engagement runs the same four phases. Below is what each one costs you in time, what you get out of it, and — because no install ever goes exactly to plan — what tends to go wrong.

01

Audit

2–5 days on site

On site

We walk the floor with your maintenance head, open panels, test comms ports, check camera streams, and map where data can come from. Nothing is connected yet; nothing stops.

You receive

  • Machine-by-machine connectivity map
  • Camera coverage and stream-quality report
  • Scoped proposal with fixed pricing per point
  • Risk register: what we found that worries us

Who's involved

Our controls engineer + your maintenance lead. Two hours of a plant manager's time.

What could go wrong

Sometimes drawings do not match the panel, or a PLC's comms port turns out dead. The audit exists to find this before you have committed money — occasionally it grows the scope, and we say so in the proposal rather than discovering it mid-install.

02

Retrofit

2–6 weeks, staged

On site

Gateways, meters, sensors, and scanners go in — machine by machine, in planned windows between shifts. Panel work follows LOTO with your electrician present. Each machine is verified reporting before we move to the next.

You receive

  • Installed and labelled hardware per the scope
  • As-built wiring drawings and tag lists
  • Per-machine signal verification record
  • Photos of every panel before and after

Who's involved

Our install team of 2–3 + your electrician for isolations. Production continues.

What could go wrong

A machine scheduled for a window sometimes cannot be spared on the day. We plan the sequence so any machine can be skipped and picked up in the next window without idle days billed to you.

03

Integrate

1–3 weeks, mostly remote

On site

Data starts flowing into dashboards. We configure shift calendars, reason codes, alert rules, and user roles with your supervisors — the people who will actually use it — then run two weeks in parallel with your existing registers.

You receive

  • Live dashboards configured to your shifts and lines
  • Alert rules agreed and tested with recipients
  • User accounts and role matrix
  • Parallel-run comparison against manual records

Who's involved

Our software engineer + your supervisors, roughly an hour a day during setup.

What could go wrong

The parallel run usually exposes disagreements between what the machines report and what the registers said. That argument is the product working. We sit in the meeting where it gets resolved.

04

Operate

Ongoing

On site

Monthly on-site reviews for the first quarter, then quarterly. We watch sensor health, tune alerts that are being ignored, and train new supervisors as your team changes.

You receive

  • Monthly data-quality and sensor-health report
  • Alert tuning and threshold reviews
  • Support: remote same-day, on-site within 48 hours in NCR
  • Quarterly review with the plant head

Who's involved

A named engineer who knows your plant. Not a ticket queue in another city.

What could go wrong

Sensors drift, cameras get repositioned, a well-meaning electrician borrows a gateway's power point. The monthly health report catches silent failures — the failure mode of every monitoring system is nobody noticing it stopped.

Your machines have been talking this whole time. Start listening.

Book a walkthrough