See which stops cost the most.
Stops by line, machine and cause from machine signals, with root cause analysis that weighs process data, asset history and similar past stops.
Stops ranked by costManufacturing
Every plant, compared on the same terms.
Machines and lines from different vendors and PLC generations, in one live view per plant and one view across plants. See where stops and energy cost you most, and act first where it pays.
Same meaning at every plantTags, states, units and timestamps aligned
Old and new PLCsSiemens S7, EtherNet/IP, Modbus, OPC UA
Lean per plantRead-only, nothing installed on PLCs
Alongside your systemsMES and ERP stay as they are
01The plant
The whole plant in one view, not one line at a time.
Built from the PLCs, SCADA and historians each plant already runs. We read the PLC directly or use OPC UA, map each plant once with your team, starting from your own tag names and descriptions, and the next plant reuses it. Nothing is installed on the PLCs.
Lines and machines
Run state, stops, counts and speeds per line, from machine signals, with good and reject counts where the PLC records them.
Site utilities
Compressed air, steam, chillers and power, next to the lines they serve.
Maintenance and spares
Work orders and spare parts from your CMMS export, on the asset they belong to.
Energy and prices
Meters, day-ahead prices and weather, so energy per unit and timing are visible.
02One meaning
Three plants, three PLCs, one answer.
Each plant keeps its own tag names, codes and units. VeldtGraph maps them once, so “which machines are running” or “parts per hour across plants” works without knowing a single tag name.
- Asset identityThe same machine, whichever system reports it.
- States and codesRun, stop and fault codes from different PLCs mapped to one set of states.
- Units and ratesParts per minute or per hour, °F or °C, converted to one unit.
- Source keptEvery value keeps its original tag, raw value and conversion.
Machine state, three plants
The same state, reported three ways, becomes one comparable reading.
| Plant tag | Raw value | Mapping | Common value |
|---|---|---|---|
| L3_STATE | 3 | code 3 = running | Running |
| Filler.RunBit | TRUE | run bit | Running |
| Mode | AUTO | AUTO + speed > 0 | Running |
| Output_PPM | 5.2 | × 60 | 312 parts/h |
Filler, Line 3Plants A, B and C
Running 312 parts/h
03One record
Everything about one machine, on one record.
Operations, maintenance, security, energy and emissions on the machine itself, and the same record for every similar machine at every plant.
One machine, the whole picture
Example: the Line 3 filler at Plant 2, as one asset record.
| View | What you see |
|---|---|
| Operations | Running, 312 parts/h, 2 short stops this shift |
| Maintenance | Drive motor vibration rising, spare motor in stock |
| Security | No unexpected connections to its controller |
| Energy & emissions | 0.06 kWh and 0.03 kg CO₂e per part, from the Line 3 power meter |
| Price | €84/MWh day-ahead, cheaper hours tonight |
| Value at risk | ≈ €18k over 30 days |
Compared across plantsSame record for every filler, at every plant
Illustrative values.
04Decisions
From signal to a decision worth making.
Every recommendation shows the data behind it and its value in euros. Your team approves every action.
Catch a failing motor or compressor early.
Health scores and failure models per failure mode (bearings, seals, electrical, thermal) show remaining useful life and the cost-optimal time to replace.
Replace on time, not too early or too lateSpot drift before it turns into scrap.
Process values watched with SPC control limits and against the same process in other plants.
Drift flagged with the data behind itSee energy per unit, by line and plant.
Meters joined to output, so the plants and lines that use the most energy per unit stand out.
Energy waste ranked by costRun flexible loads in cheaper hours.
Plan compressors, chillers and other flexible loads around day-ahead prices, within production limits.
Load plans weighed against priceRoll out what works.
A fix that worked in one plant is checked against every similar machine in the others before you roll it out.
One benchmark across every plant05Security
Know which line a security alert can stop.
Passive monitoring of your plant networks, with every alert joined to the machine and line it can affect.
The platform knows
PLC-08↓Plant 2, Line 3↓Filler↓Running, 312 parts/h
The network shows
PLC-08↓Unexpected communication↓EtherNet/IP, port 44818↓First time on this segment
What your team sees
The Line 3 filler controller is talking to an unexpected host.Operations and security see the same alert, on the same machine.
A passive sensor on a mirror port or TAP per plant network, with no software on PLCs. Data leaves the plant outbound only, encrypted. Designed to support operators’ IEC 62443 and NIS2 programs.
06Energy and emissions
Energy and emissions per line, from the meters.
Gas, power and steam meters become energy per unit and Scope 1 and 2 emissions, traced to the tag that produced them. Useful for energy audits, ISO 50001 and Scope 1 and 2 reporting.
- MeasuredActivity data read from meters and PLCs, not estimated from bills.
- Per line and per unitEnergy and CO₂e joined to what each line produced.
- TracedEvery figure links back through factor, method and source tag.
From meter to CO₂e
Example: a curing oven’s gas meter, from the source tag to the reported figure.
| Step | What happens |
|---|---|
| Source tag | FT-210 oven gas flow, read from the PLC |
| Activity data | Volume, unit and timestamp aligned by the platform |
| Per unit | Gas per part, joined to Line 2 output |
| Result | CO₂e by machine, line, plant and period |
OVEN-02, current shiftScope 1, traced to FT-210
0.42 kg CO₂e per part
Illustrative values.
07Fits what you run
Alongside your MES and ERP, not instead of them.
VeldtGraph reads what your machines already report and adds the business context you choose to share.
Machines
Read-only from the PLC.
Siemens S7, EtherNet/IP, Modbus TCP and OPC UA. Nothing installed on the PLCs. S7 PLCs are read over S7 communication on the plant Ethernet, so no PROFINET device is added.
Orders and products
Product and order context.
From an export you already have, agreed with you, so output and stops show which product and order they belong to.
Where it runs
Edge, on-premises or cloud.
Outbound-only connections, wherever it runs.
Start with one plant and one line.
Then roll the same definitions out across your plants.
A focused first engagement, measured against your own baseline.
- We map the line’s PLCs, tags and data with your team.
- We agree one outcome and a baseline, for example stop time on one line or energy per unit.
- We connect read-only, check every signal against your HMI, then add the next lines and plants.
First results in 2 to 4 weeks. You decide on rollout after you see them.