TJL
Elm

Qareen · Investor presentation

The operating intelligence layer for Saudi factories.

Qareen connects a factory's meters, machines and records into one trusted model. It then forecasts and optimises how the plant runs, for lower energy cost, lower emissions and longer asset life.

TJL · October 2026

Illustration with example values

Industry 4.0 in the Kingdom

The Kingdom is moving its factories to Industry 4.0.

Vision 2030 and the National Industrial Strategy set the direction. The Ministry of Industry's Future Factories Program moves existing plants from manual work to automated, data-driven operation.

4,000factories selected for the Future Factories Program, about 30% of the Kingdom's factories at launchMinistry of Industry, 2022
SIRIthe Smart Industry Readiness Index, the scale each factory is assessed onFuture Factories Program
260factories in the program's second wave of readiness evaluationsMinistry of Industry, 2023
TodayIllustration. Shares are not program data.

Level 1Manual

Bills, spreadsheets, paper logs.

Qareen entry: energy visibility from 12 months of bills

Level 2Metered

Smart meters or SCADA tags, little analysis.

Qareen entry: meter to dashboard, peak alerts, loss map

Level 3Integrated

SCADA, ERP and maintenance data, kept apart.

Qareen entry: factory model, energy-aware scheduling

Level 4Optimised

Real-time data, teams ready to optimise.

Qareen entry: AI optimisation, predict and verify

Sources: Arab News, "Saudi Arabia launches program to automate 4,000 factories" (2022) and "Saudi Arabia to evaluate 260 factories in modernization push" (2023) · Ministry of Industry and Mineral Resources. Levels follow the Smart Industry Readiness Index (SIRI) used in the program's assessments.

The market

14,000+ factories. 4,000 already on the path to Industry 4.0.

14,000+factories in the KingdomMinistry of Industry · industry.sa
4,000of them in the Future Factories ProgramMinistry of Industry
23 halalah/kWhindustrial electricity tariff on the distribution and transmission networks from 2027SERA

First segment: energy-intensive plants. Cement, metals, plastics, chemicals and food, where energy is the largest cost a plant can control.

4,000 of 14,000+ factories already in the program

Factories in the Kingdom Future Factories Program 1 dot = 100 factories

Sources: Ministry of Industry and Mineral Resources (industry.sa) · Saudi Electricity Regulatory Authority (SERA), consumption tariffs.

The problem

Factories pay the bill without seeing where it goes.

Day 31Time
30days from the cause to its cost
High bills, low visibilityPeak-demand and power-factor penaltiesReactive maintenance, then downtimeSchedules that ignore energy cost

Each system works on its own. The meter, the SCADA, the shift log, the maintenance system and the bill never meet, so the cause and its cost are found a month apart, if they are found at all.

Illustration with example values

The platform

One stack, from the meter to the decision.

Back to the plant:SchedulesSet pointsWork ordersReports
6

Ask

Generative and agentic AI: questions answered from the plant's own data, reports drafted, actions proposed.

Next, with Elm
5

Apply

Management dashboards, line screens, reports and recommended actions.

Running in plants
4

Forecast and optimise

Load, tariff and output forecasts. Schedules solved across seven problem domains.

Built and validated
3

Model

The digital twin: assets, processes and constraints, with physics where it matters.

Built and validated
2

Trust

Every reading mapped, converted, quality-scored and audited.

Running in plants
1

Connect

PLCs, SCADA, meters, historians, ERP and bills, read through their own interfaces.

Running in plants
MetersPLC · SCADAHistoriansERPCMMSBills

TJL built and owns every layer. A new factory is configured, not rewritten: its own assets, tags and sources run on the same code.

Traditional dashboardQareen
Reads sensorsSimulates the plant
Serves IT and analyticsServes operations and finance
MonitorsForecasts and optimises
Reports the pastChanges the profit line

What a factory gains

Five gains, measured on the factory's own data.

Energy cost

Load moved out of peak windows, demand held down, waste found machine by machine.

Savings measured against a signed baseline and checked on the bill

CO₂ emissions

Emissions computed from metered energy, per product and per line, ready to report.

Solar, storage and tariff scenarios tested on the model before any spend

Asset lifespan

Abnormal behaviour caught early, before it turns into wear. Readings, stoppages and work orders share one history per asset.

Alarms that need a condition to persist, not one bad reading

Compliance

Every figure traces back to the reading it came from. Hash-chained audit trail and role-based access.

Evidence for energy-efficiency programmes and ISO 50001 energy reviews

Big data management

Meters, PLCs, historians, ERP and bills in one model. Gaps stay gaps, never zeros.

One quality score across five dimensions

Onboarding

From a plant's own spreadsheets to live data, with no new code.

1Define the plantHierarchy, equipment and tags, read as one workbook
XLSXplant_structure.xlsx
HierarchyEquipmentTags
UOM read as unitTag Name read as source tag
Plant model
    2Connect its dataMappings proposed, tested on real data, then switched on
    OPC UAMQTTREST APIFile uploadUtility bill
    Read2,880
    Stored2,874
    Rejected6
    Duplicates0
    No codeA new plant brings its own machines, tag names and units. The application does not change.
    Proposals, not guessesEach suggested mapping shows its confidence and reason. A person accepts it.
    Tested before liveA mapping goes live only after it has produced a reading from the plant's own data.
    Same IDsWhen a pilot moves from exported files to the live connection, every signal keeps its ID.
    Illustration with example values

    Digital twin

    A running model of the process, in physics and in 3D.

    Qareen builds each process model from the equipment vendor's specifications and the plant's own data, then runs it. The physics answers how long, how much and what if. The 3D view shows the line at work.

    • What ifA schedule or a setting is tried on the model before the plant.
    • Per runDuration, energy and cost of every batch, before it starts.
    • One IDEach machine in 3D carries the same ID as its data and alarms.
    • NextEvery motion in 3D bound to live plant signals.
    Loading the 3D model…
    Temperature Melt fraction Power
    Time90min
    Temperature1,620°C
    Melt fraction100%
    Power0.0MW
    Energy10.6MWh
    Per tonne532kWh/t

    Example models from a reference deployment, with example values

    Forecast and optimise

    Forecast tomorrow's load. Solve the schedule before it runs.

    As planned
    Forecast load Forecast range Contract demand Peak window
    Peak demand26.4MW
    Energy in the peak window41MWh
    Batches delivered12/ 12

    Illustration with example values. Tariff windows differ by site.

    The engine takes a goal and the limits that must hold. It finds a plan, shows what it is worth, and records who approved it.

    • GoalsCut electricity cost, meet every due date, finish sooner, or reduce peak demand
    • LimitsMachine capacity, crews, due dates, contract demand
    • ScenariosSolar, storage and tariff changes tested on the model before any spend
    • EngineSeven problem domains, five solver backends
    Built and validated on a reference deployment

    Agentic AI

    Ask the factory. Get the answer and its source.

    Plain-language questions answered from the plant's own trusted data. Each figure names the readings and records it came from, so anyone can check it.

    Where it standsQuestions and charts from production data run daily on DigiTex lines. Agents that draft reports and propose actions are what we build with Elm.
    Illustration with example values
    Qareen assistantConnected to plant data

    Track record

    Proven on real plants in three countries.

    Rabigh Solb Factory
    Delivered · TRL 4 to 5

    Rabigh Solb Factory

    Steel mini-mill · Rabigh, Saudi Arabia

    Factory onboarded, site visit, process simulator, optimisation engine and electricity baseline.

    163 GWha year, baseline from the plant's bills and load
    8.9%modelled saving from optimisation alone
    Annual electricity cost by scenario, M SAR
    Baseline37.9
    Optimised34.5
    + 30 MWp PV22.4
    + PV, storage16.9
    DigiTex line screen on sewing line E9
    In operation · TRL 7 to 8

    DigiTex

    Garment production · Tunisia
    −56%defects per 100 pieces, week 1 to week 6
    4 minincident response, down from about 12
    +5 to +10efficiency points in twelve weeks
    8 in 8lines connected in eight weeks
    SPL Industries
    Delivered · verified baseline

    SPL Industries

    Textile dyeing · Faridabad, India
    3.13 GWha year, from 15 metered billing cycles
    6savings levers, each verifiable on the bill
    Annual value by scenario, share of electricity spend
    Conservative5.4%
    Base13.3%
    Stretch21.1%
    TALCO
    NADEC

    DigiTex was developed jointly by TJL and AdvantryX and runs in Tunisia under a local brand; its figures are those recorded by the installed system. SPL scenarios are planning figures from the verified baseline (TJL Qareen Savings Pre-Assessment, August 2026), not achieved savings. Rabigh figures are modelled on the plant's own bills and a year of measured load: site-specific, not a guarantee. Readiness levels assessed by TJL on the standard nine-level scale.

    Working software

    Built, running, and on screen today.

    The platform as it runs today: the plant dashboard, data quality, process simulation and the optimisation engine.

      Screens from a reference deployment with emulated data sources, not a live production feed.

      Qareen plant dashboard Qareen data-quality centre Qareen process simulator Qareen optimisation workspace

      Target factories for full pilots

      Five Saudi plants where Qareen's models apply from day one.

      Cement
      Yamama Cement

      Yamama Cement

      Al-Kharj, Riyadh region

      Up to 20,000 t a day. Third clinker line in commercial production in 2026. First for electricity performance in the 2025 FCI award.

      Pilot: kiln and mill energy per tonne, mill scheduling against demand

      Cement
      Arabian Cement

      Arabian Cement

      Rabigh plant · Jeddah head office

      The Kingdom's first cement company, founded 1955.

      Pilot: specific power per tonne of cement, CO₂ per tonne of clinker

      Steel
      AbdulKarim Al-Rajhi Steel

      AbdulKarim Al-Rajhi Steel

      Riyadh

      Rebar to SASO and ASTM A615, melted and thermomechanically rolled.

      Pilot: furnace model and melt scheduling, ready to apply

      Steel
      Nuor Steel

      Nuor Steel

      Saudi Arabia

      Billets and TMT rebar. Melt shop since 2014, rolling mill since 2020.

      Pilot: melt-shop energy per tonne, peak demand, rolling-mill load

      Plastics
      AlRajhi Plastic

      AlRajhi Plastic

      Al Manakh, Riyadh

      Founded 1977. Injection moulding, films, raffia and greenhouse rolls.

      Pilot: energy per kilogram, machine OEE, chillers and compressed air

      Proposed shortlist for discussion with Elm, from public information only. Sources: yamamacement.com · arabiancement.com.sa · akalrajhisteel.sa · nuorsteel.com · alrajhiplastic.com · Global Cement and Argaam on Yamama's relocated line.

      The pilot

      A 90-day pilot proves it on the factory's own data.

      Success means a signed baseline, quantified savings, and the requirements to roll out to the next site.

      1Weeks 1 to 2

      Connect

      Meters, PLCs and ERP into one clean stream.

      Weeks 1 to 2 · data connected and quality-scored

      2Weeks 3 to 6

      Model

      The virtual factory, calibrated to the site.

      Weeks 3 to 6 · a calibrated baseline

      3Weeks 7 to 10

      Simulate

      Tariff, load and schedule scenarios tested on the model.

      Weeks 7 to 10 · ranked savings levers

      4Weeks 11 to 13

      Prove

      Changes applied and measured against the baseline.

      Weeks 11 to 13 · a signed savings report

      Week 13Prove: a signed savings report
      What the factory provides 12 months of billsMeter or SCADA exportsEquipment list and shiftsProduction logsOne named contact

      What we need from Elm

      Six things that take Qareen from pilots to the program.

      TJLElm

      Developers

      A joint engineering team to run pilots in parallel and harden the platform for scale.

      Agentic AI

      Elm's AI capability on Qareen's trusted data: agents that draft reports, explain variances and propose actions.

      Hosting

      Hosting in the Kingdom, with the security and uptime that industrial clients expect.

      The Elm trademark

      Qareen offered with the Elm name, opening doors to industrial groups and government programmes.

      Pilot funding

      Full pilots at the target factories, each ending in a signed baseline and verified savings.

      Growth capital

      For the team, the product and market entry across the Future Factories base.

      What Elm gets

      1. 1

        Proof

        Paid pilots validate savings on real factory data.

      2. 2

        Platform

        An annual subscription per site, recurring across factories.

      3. 3

        Upside

        Shared savings tie revenue to measured client value.

      Plus a built, validated stack and reference deployments in three countries. Amounts to be agreed together.

      TJLElm

      TJL intelligence. Elm scale.

      Together, powering a smarter Saudi industry.

      future@tjl-ksa.org+966 59 051 0662www.tjl-ksa.org