DieBot V2 replaces the highest-risk, highest-variance manual task on the Wisconsin Centrifugal floor with a recipe-controlled automated cell — and converts 9–12 minute mold cleanings into 30-second cycles.
An operator stands over a die that is actively spinning, reaches in with a wire brush on a stick, and clears casting residue from the cavity. Wash is applied next, by hand, with a manual siphon gun — close range, open machine, elevated temperature.
Two operations, both required every cycle, both performed by hand at close proximity to rotating equipment running at elevated temperature. The result is a process that is simultaneously high-risk from a safety standpoint and inherently variable from a quality standpoint. Spray distance, angle, overlap, travel speed, temperature judgment, and coating thickness all vary by operator — and can vary within the same mold.
Manual wire-brushing inside a spinning die. Repeated every casting cycle.
Coating failures from inconsistent operator application. Direct quality and scrap impact.
Wire-brush cleaning + manual wash application. Both at close range to spinning equipment.
DieBot V2 converts die preparation from an operator-dependent process into a recipe-controlled sequence with measurable, auditable outputs. The unit performs all three between-cast operations under program control.
Motorized nozzle carrier descends into the die cavity at operator-tunable feed rates with independent high/low depth setpoints. Consistent, repeatable residue removal across the full casting surface.
Pneumatic Venturi pulls contaminated wash water out of the cavity in-process, before evaporation can leave residue behind. No mechanical pumps in the wet path. Self-draining, low-maintenance.
Programmable wash delivery lays down a consistent, repeatable film of die release. Optional integration with infrared cameras for closed-loop temperature-modulated application.
Two interchangeable modes of operation. Programmable Automated Mode runs die-specific recipes with parameters tunable per die. Augmented Manual Mode permits the operator to adjust and override in real time via sealed push-buttons and rotary speed controls without exiting the work cell. The full control package is IP67-rated for the splash-prone, elevated-temperature environment immediately adjacent to the die.
This is a working simulation of the DieBot V2 control interface. Set a target depth, choose a spray mode, and run a cycle. Every parameter shown corresponds to a real, configurable setpoint on the deployed unit.
Every interaction in this simulation maps to a real configurable parameter on the V2 unit. When deployed, recipes are saved per die and recalled by selecting a job from the unit's operator interface.
Every component is CAD-modeled, dimensioned to tolerance, and built around an IP67-rated control core. The drawings below are derived directly from the deployed unit's engineering files.
| 01 | Stepper Drive NEMA-frame · belt-coupled · top-mounted |
| 02 | Linear Rail 1300mm extrusion · ±0.05mm repeatability |
| 03 | Control Enclosure IP67 · WiFi/BT controller · sealed switchgear |
| 04 | Carriage Block Twin linear bearings · belt-driven Y travel |
| 05 | A-Frame Arm 75° diagonal compression · spray reaction load path |
| 06 | 2" Ø Mount Pin Locator pin into pre-drilled platform hole |
| 07 | Spray Head Rotating HP nozzle + Venturi wash applicator |
| — | Enclosure IP67 · sealed against dust + jet water |
| — | Controller Next-gen core · WiFi + Bluetooth onboard |
| — | Temperature Rating Elevated-temp rated for foundry-floor adjacency |
| — | Switchgear IP67 limit switches · sealed push-buttons |
| — | Speed Control IP67 rotary interface |
| — | E-Stop Mushroom-head · twist-release · side-mounted |
| Y-axis travel | Up to 59" stroke · stepper-driven · ±0.05mm rail repeatability |
| X-axis positioning | Manual lever-arm with 2" Ø locator pin · pre-drilled platform mount holes · automation-ready |
| Control core | Next-gen controller · onboard WiFi + Bluetooth · elevated-temperature rated · IP67 enclosure |
| Switchgear | IP67 limit switches · IP67 speed-control interface · sealed push-buttons throughout |
| Operating modes | Programmable Automated · Augmented Manual |
| Stage 1 — Cleaning | High-pressure water-jet · rotating conical tip · figure-8 spray pattern |
| Stage 2 — Evacuation | Venturi pneumatic vacuum · no mechanical pumps in wet path · self-draining |
| Stage 3 — Wash application | Programmable wash delivery · optional IR-camera temperature-modulated control |
| Mounting | Single-leg 2" Ø locator pin into pre-drilled operator-platform holes · no modifications to die or pit |
| Utilities required from facility | Compressed air · electrical power |
| Utilities provided in scope | High-pressure washer · fittings · plumbing · integration labor |
| Lead time | 4–6 weeks from PO to operational install |
| Service · Spare parts | Kuehl Industrial Services · stocked in Dane, WI |
| IP protection | Patents pending |
Six years of wash-related scrap from Wisconsin Centrifugal — 108 castings, $1.02M in scrap cost — show what manual die preparation has cost the facility. DieBot V2 is engineered to address the specific failure modes in that dataset.
Wash-related defect scrap costs at Wisconsin Centrifugal. Scrap only — castings reworked and saved are excluded from the totals below, so true wash-related cost is meaningfully higher than these figures.
Defect code B02 — "casting burned into die" alone accounts for $832,453 of the $1.02M wash-related scrap across six years. Burn-in is caused by inadequate or uneven wash application — exactly the failure mode DieBot's recipe-controlled wash delivery is engineered to eliminate.
Die-related safety incidents at Wisconsin Centrifugal, valued using OSHA's SafetyPays calculator at the conservative 1.1× indirect-cost multiplier. Manufacturing industry standard runs 3–10×, so true company cost is materially higher than the figures below.
DieBot removes the operator from direct exposure during the two highest-incident-frequency tasks in die preparation: wire-brush cleaning of the spinning cavity and manual wash application at close range. An 80% avoidance projection reflects that DieBot eliminates operator contact during the specific tasks that generate the majority of these events.
Incident valuations use OSHA's SafetyPays estimator (osha.gov/safetypays/estimator) with the conservative 1.1× indirect-cost multiplier. Manufacturing industry standard runs 3–10×, meaning the true cost of the ten incidents above is likely in the $2.0M–$6.4M range.
Performance metrics below reflect trial testing of the DieBot V2 platform against the current manual process. The 70% wash-defect reduction is a projected outcome — calculated as 85% of wash failures attributable to controllable process variation × 80% expected solution effectiveness — and has not yet been validated against MetalTek production samples.
9–12 min manual cleaning → 30 sec DieBot cycle. Per mold. Every cycle.
From 116 min to 99 min per 3-cast cycle. Measured in trial testing.
1.55 → 1.82 castings per hour. Same cell, no additional headcount.
Additional castings per shift from the same three-machine cell.
Engineering projection from recipe-controlled wash application. Methodology: 85% controllable × 80% effectiveness.
Combined cycle-time + defect-reduction improvement per 8-hour shift.
HP cleaning captures particulate at source. Venturi vacuum extracts before evaporation. Measured air-quality improvement facility-wide.
MetalTek's actual avg die-related incident cost, 2024–2026, per OSHA SafetyPays estimator at conservative 1.1× multiplier.
No manual operator contact with the spinning die during the cleaning or wash cycle.
Defaults reflect Wisconsin Centrifugal's actual historical performance: ~5,000 castings/year Bay 5 volume, 6-year wash-scrap averages, 2.5-year die-related incident baseline ($279,824/yr at OSHA conservative valuation). Move any input to test scenarios.
Phase 2 ($1M) payback: —
Phase 3 ($4.2M) payback: —
Reading the math: Added throughput is shown as top-line revenue at Bay 5's average casting price — not profit. Profit contribution equals this figure minus incremental production cost (typically $2K–$4K per casting) plus any variable operating expense. Throughput conversion assumes 50% of new capacity converts to sold revenue; on demand-strong programs (military, aerospace) conversion approaches 100%. Wash-scrap recovery excludes rework cost (not in the source dataset). Safety savings use OSHA's conservative 1.1× indirect-cost multiplier; industry-standard 3–10× would yield materially higher figures.
The defensive case alone — wash-scrap prevention and safety incident avoidance — is grounded in six years of scrap data and 2.5 years of incident data from your own facility. That case doesn't require Bay 5 to sell an additional casting to be true.
Bay 5 contains four vertical centrifugal dies. Phase 2 deploys one V2 DieBot serving all four — the operator moves the unit between dies between casts. Phase 3 adds three more units, one per die, eliminating movement and enabling parallel operation.
Recognition of engineering development to date. Validated proof of concept already in operation at the demonstration cell. Sign-and-approve to release the first V2 unit into Bay 5.
40 / 40 / 20 schedule over the 4–6 week delivery window. Each milestone payment due upon successful demo of that milestone's deliverables.
One V2 DieBot unit, high-pressure washers, fittings, integration labor. KIS provides all utility integration; MetalTek provides air and electrical service points.
Single DieBot unit is moved between Bay 5's four vertical dies between casts. Operator role shifts from in-cell laborer to supervisor + unit positioning.
Three additional V2 DieBot units (four total — one per die), parallel operation infrastructure, complete Bay 5 process-control buildout.
Each die has a dedicated DieBot. Zero between-die movement. All four dies serviced in parallel. Bay 5 becomes a fully recipe-controlled, instrumented production environment.
Three issues we expect will come up. Addressed directly.
DieBot deployment maps cleanly to the IRS Four-Part Test for federal R&D credit eligibility. A portion of MetalTek's Phase 1 investment may be recoverable as tax credit, materially improving the effective cost of the engagement.
DieBot V2 deployment activities at Wisconsin Centrifugal would be evaluated against each criterion below. We've mapped the work to the test:
Based on mechanical and electrical engineering, control systems, and pneumatic process design.
New process development. Improvements in quality, durability, cost reduction, and performance.
Methodology, design, and capability uncertainty addressed through iterative engineering development.
V1 → V2 iteration with trial testing, hypothesis refinement, and measured performance evaluation.
MetalTek's tax advisor (or Baker Tilly's R&D credit practice) can confirm specific eligibility and quantify the credit. KIS will support documentation of the qualifying activities.
Approve the $300,000 Phase 1 contract to release the first V2 unit into production and commit forward to the $1.0M Phase 2 Bay 5 production deployment. Phase 1 delivery in 4–6 weeks from PO. Payment 40/40/20 against milestone demonstrations.