Autonomy where your worst number lives
Treadum is deployed process by process. Pick the step that is costing you scrap, rate, energy or sleep — mixing, extrusion, building, curing, inspection or takt — and put the loop there first.
Segments we build for
Six places to start
Each has its own baseline, its own wedge metric and its own path to the rest of the line.
Mixing and compounding
Batch-to-batch rheology and dispersion consistency, mixer energy per kilogram and drift detection before the compound reaches extrusion.
Wedge: rheology varianceExtrusion and calendering
Tread and sidewall gauge, width and profile plus fabric and steel-cord ply tension held through speed and temperature changes.
Wedge: gauge scrapTire building and splice
Component placement, splice overlap and open-splice risk, bead seating and drum tension verified per component.
Wedge: splice defect rateCuring and vulcanization
Per-tire cure-state estimation, adaptive dwell, mold allocation and recoverable over-cure energy.
Wedge: cure energyInspection and uniformity
Vision, X-ray and shearography fusion plus force-variation prediction so rejects are prevented rather than sorted.
Wedge: escape and reject rateYield and takt
Scrap recovery, WIP movement, mold and press sequencing and line balancing across the whole plant.
Wedge: takt attainmentWhat Treadum looks like from your chair
The buying committee is a plant director, a process engineer and a quality engineer, plus IT/OT and safety as approvers.
Rate, scrap and safety in one place
You own production rate, quality, scrap and safety. Treadum gives you a live view of all four with the causes attached, and a rollout model that never puts the schedule at risk to get there.
- One console for yield, scrap, holds, energy and takt
- Shadow-mode baselining before anything changes
- Expansion priced per cell so ROI is provable before scale
Your process window, continuously defended
You own compounding, building and curing recipes and the process windows around them. Treadum watches every batch and every tire against those windows and proposes corrections you can accept, modify or reject.
- Every proposal carries evidence and citations
- Overrides captured as structured training data
- Twin scenarios to test a recipe change without burning production time
Evidence without the archaeology
You own conformance, defect and uniformity performance and tire genealogy. Treadum links every tire to its inputs and every action to its justification, so non-conformance and audit work stops being a manual reconstruction.
- IATF 16949, ISO 9001, DOT and UNECE R30 and R117 evidence needs
- Linked inspection imagery on the tire record
- Recall scoping as a query, not a project
Nothing leaves the segment you do not allow
You own network segmentation, identity and the risk register. Treadum runs inside the OT network, writes only through explicitly exposed setpoints and can run fully on-premises with no outbound plant data.
- Edge-resident inference, no cloud dependency for line decisions
- SSO, SAML, SCIM and RBAC with per-station scopes
- On-premises option, signed releases, pinned versions
zones:
ot:
nodes: ["edge-01", "edge-02", "edge-03"]
egress: "none" # on-prem mode
writes: "envelope-only"
dmz:
sync: "model-bundles, signed"
direction: "inbound-only"
it:
identity: "saml + scim"
reporting: "aggregates"Where Treadum works — and where it does not
You have volume, systems and a mandate
The strongest sites have enough production volume for clear ROI, existing mixing, building, curing and inspection systems with API access, an executive mandate to raise yield or cut scrap and energy, and willingness to run a paid pilot against a defined success metric.
- $200M to $40B revenue tire or rubber-goods manufacturer
- At least one cell or line where 1–3 points of first-pass yield is material
- MES, historian and inspection systems reachable under IT/OT controls
- An economic buyer who owns a hard production metric
We would rather say so early
Treadum is a system of action, not a horizontal dashboard. Sites that cannot integrate their process and inspection systems under IP and security controls, or that want reporting without any path to control, will not see the return that justifies the price.
- Tiny retreaders or rubber shops with no primary manufacturing scope
- Sites unwilling to integrate mixing, building, curing or inspection systems
- Buyers seeking a generic dashboard rather than closed-loop control
What the wedge is supposed to move
Baselines are captured in shadow mode before autonomy is activated, so the delta is not a story.
First-pass yield
A 1–3 point lift on a single high-volume line is usually enough to justify Cell pricing on its own. [ASPIRATIONAL]
Scrap and rework
Target 5–15% reduction by catching splice, placement and cure faults at the station instead of at final inspection. [ASPIRATIONAL]
Cure energy
3–8% reduction by curing to state rather than to a fixed recipe with a safety pad. [ASPIRATIONAL]
Uniformity pass rate
Force-variation rejects prevented upstream rather than sorted downstream.
Escape exposure
Fewer safety-critical field escapes, and far faster scoping when one is suspected.
How a solution deployment runs
Walk the line, pick the wedge
We map the process, agree the metric and confirm which systems expose what.
Instrument and measure
Connectors land, streams align, and the pre-Treadum baseline is captured across shifts, constructions and compounds.
Predict without writing
Accuracy, agreement and false-hold rate measured against your engineers and downstream results.
Engineers commit changes
The wedge metric starts moving, and override reasons become the training set.
Autonomy inside the envelope
Low-risk actions run automatically; the next cell or line is scoped.
Process leaders on the wedge approach
“We had three people watching X-ray images and still shipped uniformity rejects. Treadum flagged the cord shift at the building drum, not four hours later at final inspection.”
“The cure agent found 40 seconds of margin on a construction we had run the same way for eleven years. It proved it in shadow mode before it touched a press.”
“Genealogy is the part I did not expect to care about. Every tire now has a linked record from batch to grade, and audit prep went from weeks to an afternoon.”
[PLACEHOLDER] Design-partner quotes are illustrative until pilot references are published.
Why this category is moving now
Scoping a deployment
By finding the intersection of a hard metric your economic buyer owns, a station with enough signal to model, and a change that can be proven inside one quarter. Usually that is splice and placement, X-ray defect detection, cure state or uniformity prediction.
No. A single cell is a complete deployment: one wedge, one baseline, one metric. Everything else is expansion, and expansion is a configuration change.
EV load, weight and noise requirements tighten uniformity and construction tolerances, which is exactly where prediction and closed-loop control help most. Constructions are modeled per site, so new families are onboarded rather than hard-coded.
Partly. Inspection, curing and genealogy apply well; green-tire building and splice do not. We would scope it honestly rather than sell the full loop.
Put one cell on autonomy in 90 days
Pick one wedge — splice inspection, X-ray defect detection, cure-state optimization or uniformity prediction. We baseline it, run shadow mode, then graduate to bounded autonomy under a signed control envelope.
Pilots start in shadow mode. No line changes until accuracy and safety gates pass.