Manufacturing intelligence

Turn manufacturing data into better production decisions.

For steel fabricators and manufacturing job shops. Use the production data you already have to see your real capacity, find the constraint, and forecast production risk before you commit to the next order.

The raw material

Every order leaves a data trail.

Machine hours, cycle times, orders, labour, ship dates. Your plant already records them in ERP, MES, timesheets and spreadsheets. They just never meet.

The model

We turn it into a live view of capacity.

Each work center's load, measured against the hours you actually have, from receiving to packing.

The decision

See the bottleneck before it ships late.

Rebalance work, plan overtime, or quote with confidence, while there is still time to act.

Synthetic data
CAPACITY INTELLIGENCELIVE MODEL · DEMO
01 / WORK CENTER UTILIZATION 100% capacity reference
QUEUE FORECASTPeak 304 h · W49

W41 → W52 · hours waiting after each week

WORKLOAD SHARE
26%Assembly
ReceivingMachiningAssemblyQualityFinishingPacking
From the plant floor to a capacity plan. Synthetic 3D model, not client data.

Most plants don’t have a data problem. They have a decision problem. We connect the numbers you already collect to the choices you make every week: what to run, what to quote, and where to add capacity.

Where plants lose time

The data exists. It just doesn’t answer your questions yet.

Most manufacturers already track hours, orders, and output. The problem is that the numbers live in separate files and arrive too late to change the week.

  • Every department keeps its own spreadsheet.

    Which version of the backlog is the right one?

  • Weekly reports take hours of copying and pasting.

    What did we actually ship last week, by product line and by work center?

  • Capacity forecasts are educated guesses.

    Can we take this order and still ship everything else on time?

  • Bottlenecks show up as late shipments.

    Where is work waiting right now, and for how long?

  • No shared view of the plant floor.

    Are we on plan today, without walking the floor to find out?

What Insight One shows you

Four answers from the data you already have.

Your exports are mapped to one standard layout, checked for quality, then analysed. Results come as dashboards and exports, and can feed Power BI or the reporting tool you already use.

CAPACITY

Capacity & utilization

Available, used and remaining hours by work center, shift and week, instead of estimates from experience.

  • Utilization and downtime by work center
  • Remaining capacity before each due date
  • Throughput and jobs completed

CONSTRAINTS

Bottleneck & downtime intelligence

The work center that limits throughput, and what is eating its time, before it becomes overtime or a missed ship date.

  • Primary constraint with a plain-language reason
  • Downtime by cause and trend
  • Work center and shift comparison

FORECAST

Production forecasting

Expected load against available hours for the coming weeks, with a likely range instead of a single guess.

  • Weekly load versus capacity
  • Weeks at risk flagged early
  • Ship-date risk for open orders

WHAT-IF

Scenario testing

Test a change before you make it, and see where the constraint moves next.

  • An extra operator, shift or overtime
  • Reduced downtime at the constraint
  • Moving suitable work between cells

Demonstration project

A capacity cockpit for a manufacturing plant.

The kind of view a first assessment produces. Try the scenarios: every number, bar and cell updates.

CAPACITY COCKPIT · ASSEMBLY CELL · WEEKS 41–52Synthetic data
Scenario
FilterClick any bar, slice, cell or point to cross-filter
Avg available hours / week420h
Weeks over capacity5of 12
Peak load524h · W45
Hours to rebalance322h
Load vs available hours · Assembly click a week · drag to tilt
Weekly load and available hours, illustrative
Average weekly load by work center click a slice
    Work queued in front of Assembly click a week
    Utilization by work center click a cell
    Insight

    Synthetic numbers, built to show the method. A real engagement uses your orders, routings and shift hours.

    Advanced Predictive scheduling

    Know which orders will ship late, weeks before they do.

    A forecasting engine gives every open order a likely ship date and a confidence level, so late orders show up weeks before they ship. Move the capacity and watch the forecast respond.

    SHIP-DATE FORECAST · ASSEMBLY · 8 OPEN ORDERSSynthetic data
    Uncertainty
    Orders at risk3below 80% on-time
    Expected on-time77%
    Last likely shipW51median
    Capacity tested420h/week
    Ship-date probability by ordertaller = more likely · click an order
    Open orderson-time chance
    ORD-2046
    Most likely ship
    –
    Worst realistic case
    –
    On-time chance
    –

    What drives the risk

      Illustrative demonstration on synthetic data. For a customer, the forecast is built on our servers from that plant's own order history; the engine is not exposed on this page.

      How it works

      Small first step. Useful result. Then decide.

      No large up-front project. You see a working result on your own data before committing to anything bigger.

      1. STEP 1

        Data assessment

        Share one or two production exports through a private upload link, never by email. In 30 minutes we show what they can already answer about capacity and constraints.

      2. STEP 2

        First capacity view

        Your data mapped to a standard layout, a Data Quality Score, and a first view of utilization, downtime and the likely constraint.

      3. STEP 3

        Focused pilot

        One question your team cares about, answered on your real data over a few weeks, with forecasts and what-if scenarios.

      4. STEP 4

        Ongoing use

        Refresh with new exports, keep the dashboards and exports, and extend to the next question when it pays. Your data stays yours.

      See how an engagement works, step by step →

      About

      Built for the way your plant actually runs.

      Insight One is designed by an engineer with hands-on experience inside steel fabrication operations, so results reflect your real orders, work centers, shifts and ship dates, not a generic template.

      Recommendations stay practical, every number traces back to your own records, and your data stays yours.

      • ServingSteel fabricators and manufacturing job shops in Canada and the US, remote and on-site
      • EducationM.Eng., Mechanical & Manufacturing Engineering
        M.Sc., Metals Forming
        B.Sc., Materials Science & Engineering
      • CredentialsEIT, Engineers and Geoscientists BC
        P.Eng. registration in final review
      • Works withERP, MES and timesheet exports, Excel and CSV. Results export to Power BI or your own reporting tool.

      Why it works and how we engage →

      Questions

      Before you get in touch

      What is a Manufacturing Data Assessment?

      A short, no-commitment first step. You share one or two production exports (ERP, MES, timesheet or spreadsheet) through a private upload link we send you, not by email. We map them to a standard layout, score the data quality, and walk you through a first view of utilization, downtime and the likely constraint in a 30-minute session.

      Does our data need to be clean before we start?

      No. Most manufacturers start with spreadsheets, ERP exports, and timesheet reports that don’t line up perfectly. Part of the work is cleaning and connecting them, and showing you which gaps matter.

      What systems do you work with?

      Anything that can export data: ERP and production-management systems, time and attendance software, Excel files, and databases. Results come as clear dashboards and exports your team can use, and can feed Power BI or the reporting tool you already have.

      Do you work on-site?

      Yes, where it helps, for example to walk the floor and understand how work moves between stations. Most of the build and follow-up happens remotely, so we can work with plants anywhere.

      Who owns the data and the results?

      Your production data stays yours. You can export it and your results at any time, so you are never locked in. The Insight One platform, its models and its methods remain Insight One’s; for project work, deliverables and licence terms are set out in each agreement.

      Is AI or machine learning required?

      Usually not at the start. Good reporting and capacity planning solve most problems. Machine learning is recommended only when there is enough history and a clear benefit over simpler methods.

      See all questions →

      Contact

      Tell us which decision is hardest to make today.

      Tell us the decision first. Please do not email files: we agree which export helps, send you a private upload link, and in 30 minutes show what your data can already answer about capacity and constraints.