Shot-by-shot data on every press, not a clipboard reading once a shift
Willowark connects molding presses and auxiliary equipment to systems that record what actually happened on each shot, then inspects the parts those shots produced. The problem it removes is discovering a process drift at packout, several thousand parts after it started, with no record of what the machine was doing when the good parts stopped.
Where a press supports Euromap 63, Euromap 77, or OPC UA, we read setpoints, actuals, alarms, and cycle data directly. Where it does not, we hard-wire cycle and alarm contacts to get the same signals without touching the machine's control program. In-mold pressure and temperature sensors give the closest available read on part quality, and capturing the curve per shot means an alarm points at a mechanism — a short pack, a blocked gate, a hot runner zone drifting — instead of just flagging a bad part.
A molding floor is rarely one generation of equipment, and that shapes the first project. We usually start with a survey of every press to find out what each one will actually give us — a full Euromap 77 interface, an OPC UA server, an old SPI port, or nothing more than a relay contact — and design one collection approach that covers all of them rather than a different system per machine. The pilot typically goes on the presses running the tightest-tolerance or highest-scrap work, where per-shot data changes a decision fastest, and expands from there.
Reviewed
Sound familiar?
If you've said any of these, we should talk.
“We find short shots at packout, after we have already made four thousand.”
We add a check at part drop — vision, weight, or both — sized to the defect you care about, so a short shot, flash, or missing insert is caught within a few cycles. When parts carry cavity identification marks, the reject is attributed to the cavity instead of condemning the whole shot.
“The cycle time on the floor and the cycle time in our quote are different numbers.”
Automatic cycle capture per tool per press gives you actual cycle, uptime, and parts per hour as history. Quoting then works from measured performance on that specific tool and machine pairing, including the startup scrap and changeover time estimates usually forget.
“We do not really know how much regrind is in a given run.”
We integrate gravimetric blenders and dryers so actual blend ratio, dew point, and drying time are logged against the job. The record travels with the lot, which matters when a customer asks or a mechanical property complaint needs explaining.
“Mold maintenance happens when the tool breaks.”
Cycle counters follow each tool from press to press, so preventive maintenance triggers on accumulated cycles rather than the last time someone remembered. Cleanings, vent work, and ejector repairs live with the tool record, making recurring problems visible instead of anecdotal.
“Every mold change takes half a shift and we never know why one went fast and one went slow.”
We time each changeover from last good shot to first good shot, broken into the steps that make it up: tool out, tool in, water and hot runner connect, purge, and first-article approval. With that breakdown, slow changeovers point to a specific step, a specific tool, or a missing setup sheet instead of a general complaint.
How this industry actually runs
The operation as we usually find it.
Molding decides quality in a few seconds using variables nobody can see from outside the machine: melt temperature, injection velocity profile, transfer position, pack pressure and time, cooling, and screw recovery. Around that sit the practical realities — hygroscopic resin that was not dried long enough, hot runner zone faults, regrind ratios drifting because a blend was eyeballed, mold changes measured in hours of lost capacity, and tools that need maintenance scheduled on cycle counts rather than the calendar. A 32-cavity tool with one blocked gate will quietly produce short shots from the same cavity all run, and if parts drop into a common bin, nobody knows which cavity to pull.
Machine signals to the people who decide
Components:
- PLCs & sensors (counts, states, current)
- Legacy machine (dry contact / clamp)
- Edge gateway (normalize, buffer)
- Production dashboard (downtime, OEE)
- Alerts & reports (who acts, when)
Connections:
- PLCs & sensors to Edge gateway (EtherNet/IP, Modbus)
- Legacy machine to Edge gateway
- Edge gateway to Production dashboard (MQTT)
- Edge gateway to Alerts & reports
What we build
Starting projects that fit Plastics & Injection Molding.
- Press data collection over Euromap 63, Euromap 77, OPC UA, or hard-wired cycle and alarm contacts
- Cavity pressure and mold temperature monitoring with per-shot curve capture and alarm limits
- Vision or check-weigh verification at part drop for short shots, flash, sink, and missing inserts
- Per-cavity traceability using cavity identification marks so a bad cavity is identified, not guessed
- Tool life tracking with cycle-count-based preventive maintenance and full tool history
- Dryer, blender, and chiller monitoring: dew point, drying time, regrind ratio, water temperature
- Scrap and startup-waste reporting by tool, press, resin lot, operator, and shift
- Changeover tracking from last good shot to first good shot with step-level timing per tool and press
Capabilities we bring
Working in Plastics & Injection Molding?
Tell us the line.
What runs by hand, what is not connected, what you are trying to build. An engineer replies within one business day with whether and how we would approach it.
Common questions
What Plastics & Injection Molding teams ask first.
Our presses are twenty years old. Can they be monitored at all?
Yes. Older machines usually expose enough through relay contacts, an operator panel signal, or an existing SPI connection to give reliable cycle counts, alarms, and uptime. You lose the detailed setpoints a modern control publishes, but cycle, downtime, and part count carry most improvement work.
Is cavity pressure monitoring worth it for our part mix?
It pays back fastest on tight-tolerance parts, medical or automotive work with documentation requirements, and multi-cavity tools where cavity-to-cavity variation is the problem. For simple commodity parts, cycle and scrap data deliver more per dollar, and we will say so before quoting sensors.
Can vision catch cosmetic defects like sink, splay, and flow lines?
Often, with the right lighting geometry, since these are surface and subsurface effects that need grazing or structured illumination to appear at all. We answer it with your actual rejects on a bench first, because some cosmetic calls are subjective and need a documented standard before automation.
Do we need a full MES, or can press monitoring stand on its own?
Press monitoring usually stands on its own well, and for many molders it is the right first step: cycle, downtime, scrap, and alarm data per press and tool, with reporting your scheduler and quoting team can use. A full MES adds job scheduling, material tracking, and labor, which some shops need and many do not. We typically build monitoring so those pieces can be added later without replacing what is already running.
Our customers want process data with every shipment. Can that be automated?
Typically yes, once the data is being captured per shot. A shipment record can pull the process window, actual cycle data, resin lot, dryer readings, and any inspection results for the parts in that lot and package them as a report in the format your customer asks for. What it cannot do is invent data that was never collected, which is usually the real gap and the reason to instrument the press first.
Strategy. Software. Systems.
Engineering for Plastics & Injection Molding.
Describe the problem in your own words. An engineer reads it — not a sales script — and tells you plainly what it would take.