Financial Analysis — Tilden Mine¶
Mission: Bottom-up value and cost analysis for every Tilden initiative. Each card defines the formula and variables — the math structure, not necessarily the final number. Numbers are tagged by source:
workshop-confirmed,estimated, orneeds-corporate. This is the audit trail: every dollar in the business package traces back to a card here.Methodology: Value per §3 of consolidation-plan.md (5-step: metric → current state → target → delta → sensitivity). Cost per §4 (effort-based: team composition × duration × rates).
Important: Tilden is a MINE, not a steel mill. Value drivers are reagent spend, ore recovery, fleet lifecycle, pellet quality, and logistics — not throughput per heat or coil quality. Mining equipment costs are higher per unit ($12M trucks, $30M shovels, $70K tires) and the operational environment (open-pit, mobile fleet, seasonal constraints) is fundamentally different.
Last updated: 2026-04-16 (appendix preparation — no content changes, audit trail preserved as constructed)
How to Read This File¶
Initiative cards define the value formula and cost structure for each of Tilden's 53 initiatives. They are grouped by parent site project (TLD-P01..P16).
Three card types: - Type A — Anchored: Formula defined, key variables have workshop-confirmed values. Math can be partially computed now. - Type B — Structured: Formula defined, but most variables need corporate data to populate. The math is built, the inputs are TBD. - Type C — Absorbed: No standalone formula. Value captured in parent project (enabler, seed, or subsumed scope).
Project roll-ups aggregate initiative cards into site project totals.
Corporate Inquiry Table (end of file) collects all needs-corporate variables into one table for IE to take to Cleveland-Cliffs.
TLD-P01: Concentrator Feed-Forward & Ore Intelligence ★★★ FLAGSHIP¶
TLD-21: Concentrator Feed-Forward Control¶
Card Type: A — Anchored Corporate Project: new (mining-specific)
Value Analysis¶
Value Types: Reagent savings + Recovery improvement + Throughput stability Value Formula:
(reagent_annual_spend × reagent_reduction_%)
+ (recovery_improvement_% × annual_pellet_tons × pellet_value_per_ton)
+ (reactive_adjustment_hours_saved × throughput_value_per_hour)
| Variable | Value | Source | Status |
|---|---|---|---|
| reagent_annual_spend | ~$50M/yr | Site leader: "probably 50 million dollars a year on chemicals" | workshop-confirmed |
| reagent_reduction_% | 5-10% | Conservative — reactive dosing wastes on mismatched ore | estimated |
| recovery_improvement_% | 1-5% (from 70% baseline toward ~75% design benchmark, realistic upside to 80%) | Each 1% = ~77K additional tons | estimated |
| annual_pellet_tons | ~7.7M tons | Site production target | workshop-confirmed |
| pellet_value_per_ton | $100+ | Market pellet price | needs-corporate |
| reactive_adjustment_hours_saved | [TBD] | Current: days-long feedback loop → target: same-shift | needs-corporate |
| throughput_value_per_hour | [TBD] | Concentrator throughput × pellet margin | needs-corporate |
Workshop-Sourced Range: $2.5-5M/yr Confidence: High — data exists on both sides of the gap (drill data + concentrator DCS), problem articulated by 5+ stakeholders Key Quotes: "If there was some learning based on the ore quality that comes in from the mining area, what adjustments happen in the concentrator in order to proficiently process it — that'd be super beneficial." "We don't know which lever to pull at some times."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist, PM | [TBD] | Drill-to-concentrator data mapping |
| Data engineering | Data engineer (senior) | [TBD] | Modular + drill data + DCS + Pi historian integration |
| ML/AI development | ML engineer, data scientist | [TBD] | Feed-forward prediction model, ore→reagent mapping |
| Application/UX | Frontend dev | [TBD] | Per-section adjustment dashboard for control operators |
| Infrastructure | Moderate | [TBD] | Data pipeline from pit to plant, possible edge compute |
| Change management | — | [TBD] | Moderate — process engineering buy-in critical. 20%. |
TLD-06: Ore Grade Control & Blend Optimization¶
Card Type: B — Structured Corporate Project: new (mining-specific)
Value Analysis¶
Value Types: Recovery improvement + Quality consistency Value Formula:
(off_spec_pellet_rate × pellet_tons_per_year × margin_loss_per_off_spec_ton × reduction_%)
+ (recovery_improvement_% × annual_pellet_tons × pellet_value_per_ton)
| Variable | Value | Source | Status |
|---|---|---|---|
| off_spec_pellet_rate | [TBD] | Quality records | needs-corporate |
| pellet_tons_per_year | ~7.7M | Production target | workshop-confirmed |
| margin_loss_per_off_spec_ton | [TBD] | Downgrade pricing vs. spec pricing | needs-corporate |
| reduction_% | 20-40% | Improved blend = fewer off-spec | estimated |
| recovery_improvement_% | 0.5-2% | Better blend → better concentrator performance | estimated |
| pellet_value_per_ton | $100+ | Market pellet price | needs-corporate |
Workshop-Sourced Range: $3-8M/yr Confidence: Medium-High — ore variability confirmed, drill hole data exists, geological model exists Key Quote: "We sample every single drill hole... to understand what's there to help us predict how the concentrator is going to react."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, domain specialist, PM | [TBD] | Geological model integration scoping |
| Data engineering | Data engineer | [TBD] | Vulcan model + assay data + dispatch integration |
| ML/AI development | Data scientist | [TBD] | Blend optimization model |
| Application/UX | Frontend dev | [TBD] | Blend recommendation dashboard |
| Infrastructure | Moderate | [TBD] | Vulcan data access, mine planning integration |
| Change management | — | [TBD] | High — mine engineering + process engineering alignment. 25%. |
TLD-31: Stockpile Ore Distribution Modeling¶
Card Type: A — Anchored Corporate Project: new (mining-specific)
Value Analysis¶
Value Types: Reagent savings + Recovery consistency Value Formula:
per_section_reagent_variance × sections × shifts_per_year × reduction_%
+ recovery_consistency_improvement_% × annual_pellet_tons × pellet_value_per_ton
| Variable | Value | Source | Status |
|---|---|---|---|
| per_section_reagent_variance | [TBD] | Concentrator cost data by section | needs-corporate |
| sections | 12 (grouped as 2-3, 4-6, 7-9, 10-12) | Concentrator layout confirmed | workshop-confirmed |
| shifts_per_year | ~1,095 (3 shifts × 365) | 24/7 operations | workshop-confirmed |
| reduction_% | 30-50% | Knowing what ore is in each section | estimated |
| recovery_consistency_improvement_% | 0.5-1% | Reduced firefighting from section variation | estimated |
| annual_pellet_tons | ~7.7M | Production target | workshop-confirmed |
| pellet_value_per_ton | $100+ | Market pellet price | needs-corporate |
Workshop-Sourced Range: $2-5M/yr Confidence: Med-High — "We have the hardware we need." GPS, truck quality data, tripper position all confirmed. Zero hardware investment. Key Quote: "If we had some knowledge of where along the modeling, is that ore in terms of which section represents what — we may be able to come back and control."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, PM | [TBD] | Data audit: GPS/truck/tripper/quality linkage |
| Data engineering | Data engineer | [TBD] | Modular GPS + crusher timestamps + tripper position integration |
| ML/AI development | ML engineer, data scientist | [TBD] | Per-section distribution model, self-improving with response correlation |
| Application/UX | Frontend dev | [TBD] | Per-section quality prediction dashboard |
| Infrastructure | Minimal | [TBD] | Pure data integration — no new hardware |
| Change management | — | [TBD] | Low — process engineers proposed this themselves. 15%. |
Note: TLD-31 is the recommended Phase 1 entry point for TLD-P01. More tractable than full feed-forward control, delivers early value.
TLD-23: Reagent Suite Optimization¶
Card Type: A — Anchored Corporate Project: new (mining-specific)
Value Analysis¶
Value Types: Direct cost reduction Value Formula:
reagent_annual_spend × optimization_savings_%
+ recovery_improvement_from_better_dosing_% × annual_pellet_tons × pellet_value_per_ton
| Variable | Value | Source | Status |
|---|---|---|---|
| reagent_annual_spend | ~$50M/yr | Site leader confirmed | workshop-confirmed |
| optimization_savings_% | 5-10% | 1974 design vs. current ore = large gap | estimated |
| recovery_improvement_from_better_dosing_% | 0.5-2% | Better reagent response to ore variability | estimated |
| annual_pellet_tons | ~7.7M | Production target | workshop-confirmed |
| pellet_value_per_ton | $100+ | Market pellet price | needs-corporate |
Workshop-Sourced Range: $2.5-5M/yr Confidence: Medium-High — $50M/yr anchor confirmed, 1974 reagent design vs. changing ore body Key Quote: "The plant and the reagent suite was designed in 1974, based on the ore quality we were seeing when we started mining... our reagents don't react the same way as they did in 1974."
Cost Analysis¶
Bundled with TLD-21 feed-forward control. Reagent optimization is a downstream output of the same prediction model. Marginal cost.
TLD-P02: Concentrator Operations & Recovery Optimization¶
TLD-07: Concentrator AG Mill Throughput Optimization¶
Card Type: B — Structured Corporate Project: PRJ-04 (reframed)
Value Analysis¶
Value Types: Throughput gain + Energy savings Value Formula:
(throughput_improvement_% × 12_mills × current_throughput_per_mill × pellet_value_per_ton)
+ (energy_savings_from_reduced_overgrinding × energy_cost_per_kWh)
| Variable | Value | Source | Status |
|---|---|---|---|
| throughput_improvement_% | 2-5% | G2 augmentation with ML | estimated |
| 12_mills | 12 AG mills confirmed | Site walkthrough | workshop-confirmed |
| current_throughput_per_mill | [TBD] | DCS production data | needs-corporate |
| pellet_value_per_ton | $100+ | Market pellet price | needs-corporate |
| energy_savings_from_overgrinding | [TBD] | Currently overgrinding when pebble mill constrained | needs-corporate |
| energy_cost_per_kWh | [TBD] | Power contract | needs-corporate |
Workshop-Sourced Range: $2-5M/yr Confidence: Medium-High — G2 fuzzy logic exists as augmentable foundation, DCS data confirmed Key Quote: "Concentrating your primary bottleneck? Yep."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist, PM | [TBD] | G2 control system audit, DCS data mapping |
| Data engineering | Data engineer | [TBD] | DCS + G2 + Pi historian integration |
| ML/AI development | ML engineer, data scientist | [TBD] | ML optimization layer on G2 fuzzy logic |
| Application/UX | Frontend dev | [TBD] | Operator guidance dashboard |
| Infrastructure | Moderate | [TBD] | DCS integration, real-time inference |
| Change management | — | [TBD] | Moderate — SGS/G2 partnership needed. 20%. |
TLD-08: Flotation Recovery Optimization¶
Card Type: A — Anchored Corporate Project: PRJ-04 (reframed)
Value Analysis¶
Value Types: Recovery improvement (highest single-value item at Tilden) Value Formula:
recovery_improvement_% × annual_pellet_tons × pellet_value_per_ton
+ reagent_dosing_optimization_savings
| Variable | Value | Source | Status |
|---|---|---|---|
| current_recovery_rate | ~70% | Site leader: "almost like 70" | workshop-confirmed |
| target_recovery_rate | 75-80% | ~75% is the plant's own design benchmark; 80% achievable with optimized hematite flotation. (Note: 90%+ at CLF Minnesota taconite operations reflects magnetic separation of magnetite — a fundamentally different process.) | estimated |
| recovery_improvement_% | 5-10% (from ~70% toward 75-80%) | Each 1% = ~77K additional tons at $100+/ton | estimated |
| annual_pellet_tons | ~7.7M | Production target | workshop-confirmed |
| pellet_value_per_ton | $100+ | Market pellet price | needs-corporate |
| reagent_dosing_optimization_savings | [TBD] | Subset of $50M/yr reagent spend | needs-corporate |
| silica_reading_frequency | 1 per 14 min | Courier machine for both units | workshop-confirmed |
| per_line_data | None — 6 lines per unit, no per-line data | Process engineering confirmed | workshop-confirmed |
Workshop-Sourced Range: $4-10M/yr (recovery improvement from 70% toward 75-80% hematite flotation benchmark, plus reagent dosing optimization) Confidence: Medium-High — recovery gap confirmed, mechanism understood, instrumentation gap identified. Note: the ~75% design benchmark and ~80% optimized ceiling reflect hematite flotation; the 90%+ figure at CLF's Minnesota operations is from magnetic separation of magnetite, a fundamentally different process. Key Quote: "You don't know if it's just one line causing the problem, or if there's an ore change and they're all causing the problem."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist | [TBD] | Flotation circuit instrumentation audit |
| Data engineering | Data engineer | [TBD] | DCS + flotation cell + assay integration |
| ML/AI development | ML engineer, data scientist | [TBD] | Per-line recovery model, reagent optimization |
| Application/UX | Frontend dev | [TBD] | Flotation performance dashboard |
| Infrastructure | Moderate | [TBD] | Per-line instrumentation needed (capex) |
| Change management | — | [TBD] | Moderate — operator process change. 20%. |
Note: Flotation improvement requires additional per-line instrumentation investment beyond software.
TLD-22: Filter Performance Monitoring (42 Filters)¶
Card Type: A — Anchored Corporate Project: new
Value Analysis¶
Value Types: Throughput gain (directly gates bottleneck) Value Formula:
filter_degradation_events_per_year × avg_cascade_days × throughput_loss_per_day × prevention_rate
+ maintenance_efficiency_from_early_detection × labor_cost
| Variable | Value | Source | Status |
|---|---|---|---|
| filter_count | 42 | Confirmed | workshop-confirmed |
| instrumentation_cost_total | ~$125K | "$2-3K per filter + $375-500 share of AI/AO module" | workshop-confirmed |
| detection_lag_current | 2-3 days | Site leader: "don't even recognize it for two or three days" | workshop-confirmed |
| cascade_failures_before_detection | 3-4 filters | "Three or four problems pile up" | workshop-confirmed |
| filter_degradation_events_per_year | [TBD] | Maintenance/ops records | needs-corporate |
| throughput_loss_per_day_when_filter_constrained | [TBD] | Concentrator capacity reduction | needs-corporate |
| prevention_rate | 70-90% | Same-shift detection vs. 2-3 day lag | estimated |
Workshop-Sourced Range: $1-3M/yr Confidence: High — one pilot filter on DCS "works really good," cascading failure mechanism clearly described, ~$125K hardware cost quantified Key Quote: "Things can pop up and you don't even recognize it for two or three days, and you might not even recognize there's a problem until three or four problems pile up on each other."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, PM | [TBD] | Filter performance baseline, anomaly threshold design |
| Data engineering | Data engineer | [TBD] | DCS integration for 42 filters (~$125K hardware) |
| ML/AI development | ML engineer | [TBD] | Anomaly detection per filter, degradation prediction |
| Application/UX | Frontend dev | [TBD] | Filter health dashboard with alert system |
| Infrastructure | Hardware: ~$125K | [TBD] | Sensors + AI/AO modules (11 needed for 42 filters) |
| Change management | — | [TBD] | Low — ops actively wants this. 15%. |
TLD-32: Concentrator Operator Decision Support¶
Card Type: A — Anchored Corporate Project: new
Value Analysis¶
Value Types: Throughput gain + Training efficiency Value Formula:
six_bearing_events_per_year × failure_rate × tons_lost_per_failed_event × pellet_value_per_ton × improvement_%
+ operator_training_time_saved × new_operators_per_year × labor_cost
| Variable | Value | Source | Status |
|---|---|---|---|
| six_bearing_success_rate | ~75% (25% failure) | "75% of the time the system works. 25% of the time we lose tons." | workshop-confirmed |
| target_success_rate | 90-95% | Best-operator pattern propagation | estimated |
| six_bearing_events_per_year | [TBD] | DCS event logs | needs-corporate |
| tons_lost_per_failed_event | [TBD] | Concentrator throughput loss per failed intervention | needs-corporate |
| pellet_value_per_ton | $100+ | Market pellet price | needs-corporate |
| operator_training_time_saved | [TBD] | Months to competency currently | needs-corporate |
Workshop-Sourced Range: $1-3M/yr Confidence: High — failure rate quantified, operators actively requesting this tool Key Quote: "Could this software teach in the moment?"
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist, PM | [TBD] | G2 logic analysis, event pattern mapping |
| Data engineering | Data engineer | [TBD] | DCS + G2 event data integration |
| ML/AI development | ML engineer, data scientist | [TBD] | Event classifier, best-operator pattern model |
| Application/UX | Frontend dev | [TBD] | Operator recommendation display, teaching mode |
| Infrastructure | Moderate | [TBD] | Real-time inference at DCS control level |
| Change management | — | [TBD] | Moderate — trust-building with control operators. 20%. |
TLD-11: Concentrator Energy Optimization¶
Card Type: C — Absorbed Corporate Project: new Reason: Seed — limited field evidence. Energy savings captured as a component of TLD-07 (mill optimization) and TLD-08 (flotation). Value Contribution: $0.5-1M/yr estimated from reduced overgrinding — captured in TLD-P02 roll-up. Cost Contribution: One analytics module within TLD-P02 scope.
TLD-P03: Pellet Plant Quality & Control¶
TLD-34: Pellet Calcium Control Automation¶
Card Type: A — Anchored Corporate Project: new
Value Analysis¶
Value Types: Quality consistency + Labor efficiency Value Formula:
(quality_variance_from_manual_control × pellet_tons_affected × margin_loss_per_off_spec)
+ (labor_hours_saved_per_year × labor_cost_per_hour)
| Variable | Value | Source | Status |
|---|---|---|---|
| current_adjustment_frequency | Every 6 hours (human) | Process engineering: "takes a human every six hours" | workshop-confirmed |
| data_availability | "More than enough data" | Process engineering team | workshop-confirmed |
| quality_variance_from_manual_control | [TBD] | Lab quality records — calcium compliance rate | needs-corporate |
| pellet_tons_affected | [TBD] | Volume between adjustments | needs-corporate |
| margin_loss_per_off_spec | [TBD] | Downgrade penalty | needs-corporate |
| labor_hours_saved_per_year | [TBD] | Operator adjustment time x 4/day x 365 | needs-corporate |
| labor_cost_per_hour | [TBD] | Loaded operator rate | needs-corporate |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High — team explicitly called this "easy application test case," data confirmed Key Quote: "If you're looking for an easy application test case, calcium control. We're already pretty good at that. It takes a human every six hours making adjustments. We have more than enough data."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, PM | [TBD] | Control parameter mapping |
| Data engineering | Data engineer | [TBD] | DCS + lab data integration |
| ML/AI development | ML engineer | [TBD] | Predictive control model — well-defined control problem |
| Application/UX | Frontend dev | [TBD] | Operator override + monitoring interface |
| Infrastructure | Minimal | [TBD] | DCS integration exists |
| Change management | — | [TBD] | Low — team proposed this themselves. 15%. |
TLD-09: Pellet Quality Prediction¶
Card Type: B — Structured Corporate Project: PRJ-04 (reframed)
Value Analysis¶
Value Types: Quality improvement + Energy savings Value Formula:
off_spec_pellet_rate × annual_pellet_tons × margin_loss_per_off_spec_ton × reduction_%
+ energy_savings_from_optimized_firing_per_ton × annual_pellet_tons
+ explosion_incidents_per_year × cost_per_incident × prevention_rate
| Variable | Value | Source | Status |
|---|---|---|---|
| off_spec_pellet_rate | [TBD] | Quality records | needs-corporate |
| annual_pellet_tons | ~7.7M | Production target | workshop-confirmed |
| margin_loss_per_off_spec_ton | [TBD] | Downgrade pricing | needs-corporate |
| reduction_% | 20-40% | Predictive model reduces variability | estimated |
| energy_savings_per_ton | [TBD] | Optimized kiln firing profile | needs-corporate |
| explosion_incidents_per_year | [TBD] | Moisture events in preheat zone | needs-corporate |
| cost_per_incident | [TBD] | Downtime + repair + safety | needs-corporate |
Workshop-Sourced Range: $2-5M/yr Confidence: Medium-High — balling operator skill variation confirmed as key variable ("five-year-old to Picasso")
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist, PM | [TBD] | Pellet process mapping, operator skill assessment |
| Data engineering | Data engineer | [TBD] | DCS + lab + kiln instrumentation |
| ML/AI development | ML engineer, data scientist | [TBD] | Quality prediction model — multi-variable |
| Application/UX | Frontend dev | [TBD] | Balling operator guidance, kiln optimization |
| Infrastructure | Moderate | [TBD] | DCS integration, lab data access |
| Change management | — | [TBD] | Moderate — operator skill sensitivity. 20%. |
TLD-10: Kiln & Grate Temperature Optimization¶
Card Type: C — Absorbed Corporate Project: new Reason: Seed — not validated with specific stakeholders. Value captured in TLD-09 (quality prediction includes kiln optimization). Value Contribution: $0.5-2M/yr estimated from energy savings — captured in TLD-P03 roll-up. Cost Contribution: One optimization model within TLD-P03 scope.
TLD-P04: Mining Fleet PdM & Lifecycle Intelligence ★★★¶
TLD-19: Tire Management & Prediction¶
Card Type: A — Anchored (LEAD STEPPING STONE) Corporate Project: PRJ-03
Value Analysis¶
Value Types: Cost avoidance + Procurement optimization Value Formula:
(tire_annual_spend × life_extension_%)
+ (allotment_forecast_error_cost × forecast_improvement_%)
+ (catastrophic_tire_failures_per_year × cost_per_failure × prevention_rate)
| Variable | Value | Source | Status |
|---|---|---|---|
| tire_annual_spend | ~$7.5M/yr (108 tires x $70K) | Pete Austin: confirmed | workshop-confirmed |
| life_extension_% | 10-15% | Optimized front-to-rear rotation timing, duty-cycle awareness | estimated |
| allotment_forecast_error_cost | [TBD] | Premium for off-cycle orders, world tire shortage risk | needs-corporate |
| forecast_improvement_% | 20-40% | AI vs. manual annual forecasting for August Bridgestone order | estimated |
| catastrophic_tire_failures_per_year | [TBD] | Mostly road hazards, some wear-out | needs-corporate |
| cost_per_failure | $70K + downtime + potential wheel motor damage ($300K) | Workshop data | workshop-confirmed |
| front_tire_life | ~2,000 operating hours | Pete Austin confirmed | workshop-confirmed |
| tire_lifecycle | Front (2 tires, new) then Rear (4 tires, used) then run to failure | Workshop confirmed | workshop-confirmed |
| chain_cost | $100K/set | Shop visit | workshop-confirmed |
Workshop-Sourced Range: $1-3M/yr Confidence: High — monitoring infrastructure exists, detailed lifecycle data from supplier, clear annual cycle (August allotment) Key Quote: "In August we have to tell Bridgestone how many tires we're going to use next year."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, PM | [TBD] | Tire lifecycle data mapping |
| Data engineering | Data engineer | [TBD] | Tire monitoring + Modular + ELLIPS + Bridgestone portal integration |
| ML/AI development | ML engineer, data scientist | [TBD] | Per-tire RUL model, rotation optimizer, allotment forecaster |
| Application/UX | Frontend dev | [TBD] | Tire health dashboard, allotment planner |
| Infrastructure | Minimal | [TBD] | Monitoring already exists — data integration |
| Change management | — | [TBD] | Low — Pete Austin is the champion and user. 15%. |
TLD-02: Heavy Mobile Equipment PdM (Trucks & Shovels)¶
Card Type: A — Anchored Corporate Project: PRJ-03
Value Analysis¶
Value Types: Cost avoidance + Availability improvement Value Formula:
(unplanned_failures_per_year × avg_repair_cost × prevention_rate)
+ (availability_improvement_% × fleet_value × annual_operating_hours × margin_per_hour)
+ (overloading_damage_cost_per_year × reduction_%)
| Variable | Value | Source | Status |
|---|---|---|---|
| truck_count | 14 fleet (320-ton + 150-ton) | Workshop confirmed | workshop-confirmed |
| shovel_count | 4 (Komatsu P&H electric rope) | Workshop confirmed | workshop-confirmed |
| truck_value | ~$12M each | Pete Austin | workshop-confirmed |
| shovel_value | ~$27-30M each | Pete Austin | workshop-confirmed |
| engine_cost | ~$1M per engine | Workshop confirmed | workshop-confirmed |
| wheel_motor_cost | ~$300K each, 4th rebuild | Workshop confirmed | workshop-confirmed |
| current_truck_availability | ~85% | Workshop: "truck availability ~85%" | workshop-confirmed |
| current_shovel_availability | High 80s% | Workshop | workshop-confirmed |
| PM_rate | ~70/30 PM/reactive | Workshop confirmed | workshop-confirmed |
| unplanned_failures_per_year | [TBD] | ELLIPS records | needs-corporate |
| avg_repair_cost | [TBD] | ELLIPS cost data | needs-corporate |
| prevention_rate | 20-30% | Conservative H1 | estimated |
| overloading_damage_cost | [TBD] | 15% overloading impact on engines/tires | needs-corporate |
| telemetry_extraction_frequency | Every 3-4 months via laptop | Pete: "untapped data right now" | workshop-confirmed |
Workshop-Sourced Range: $3-8M/yr Confidence: High — proven at scale in global mining (Rio Tinto, BHP, Vale), data exists but fragmented Key Quote: "Either we didn't buy the subscription or it's not going to work the way we wanted to — it's untapped data right now." — Pete Austin
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist, PM | [TBD] | Multi-OEM data audit, asset selection |
| Data engineering | Data engineer | [TBD] | Cat/Komatsu + Modular + ELLIPS consolidation |
| ML/AI development | ML engineer, data scientist | [TBD] | RUL models per component, fault code normalization |
| Application/UX | Frontend dev | [TBD] | Fleet health dashboard, alert management |
| Infrastructure | Moderate | [TBD] | Automated telemetry extraction (replace manual laptop) |
| Change management | — | [TBD] | Low — Pete Austin's team actively wants this. 15%. |
TLD-46: Duty-Cycle Based Maintenance (Tons vs Hours)¶
Card Type: B — Structured Corporate Project: new (mining paradigm shift)
Value Analysis¶
Value Types: Maintenance optimization + Cost avoidance Value Formula:
(over_maintained_assets × excess_PM_cost_per_asset)
+ (under_maintained_assets × excess_failure_cost_per_asset)
+ (PM_interval_optimization_savings_% × total_fleet_maintenance_spend)
| Variable | Value | Source | Status |
|---|---|---|---|
| shovel_tonnage_disparity | #1 shovel sees 10x more tons than bottom priority | Pete Austin confirmed | workshop-confirmed |
| fuel_burn_variability | 30-60 gallons/hour depending on duty | Pete Austin confirmed | workshop-confirmed |
| engine_fuel_lifecycle | ~1.4M gallons (not flat hours) | Pete: "should be able to go 1.4 million gallons of fuel" | workshop-confirmed |
| total_fleet_maintenance_spend | [TBD] | ELLIPS + financial records | needs-corporate |
| excess_PM_cost_% | [TBD] | Over-maintenance on light-duty assets | needs-corporate |
| excess_failure_cost_% | [TBD] | Under-maintenance on heavy-duty assets | needs-corporate |
Workshop-Sourced Range: $2-5M/yr Confidence: Med-High — concept validated, partial precedent (shovel ropes already on tonnage) Key Quote: "Not every hour is equal. The number one priority unit — you're going to see trucks all the time. Bottom priority, you might see a couple trucks an hour."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, domain specialist, PM | [TBD] | Component-level duty cycle analysis |
| Data engineering | Data engineer | [TBD] | Modular tonnage + onboard fuel + ELLIPS maintenance fusion |
| ML/AI development | ML engineer, data scientist | [TBD] | Duty-cycle weighted maintenance models per component |
| Application/UX | Frontend dev | [TBD] | Fleet PM scheduling dashboard |
| Infrastructure | Moderate | [TBD] | Automated onboard data extraction needed |
| Change management | — | [TBD] | High — paradigm shift in maintenance philosophy. 25%. |
TLD-47: Fleet Capital Replacement & Lifecycle Planning¶
Card Type: B — Structured Corporate Project: new (strategic)
Value Analysis¶
Value Types: Capital optimization Value Formula:
(suboptimal_replacement_timing_cost × fleet_size)
+ (synchronized_aging_risk × fleet_value)
+ (repair_vs_replace_decision_improvement × annual_rebuild_spend)
| Variable | Value | Source | Status |
|---|---|---|---|
| truck_replacement_cost | ~$12M each | Workshop confirmed | workshop-confirmed |
| shovel_replacement_cost | ~$27-30M each | Workshop confirmed | workshop-confirmed |
| fleet_total_value | ~$280M+ (14 trucks + 4 shovels) | Calculated | estimated |
| crossover_point | ~120,000 hours (cumulative replacement > new truck) | Pete Austin | workshop-confirmed |
| wheel_motor_rebuild_trajectory | 4th rebuild, each costing more | Workshop confirmed | workshop-confirmed |
| annual_rebuild_spend | [TBD] | Financial records | needs-corporate |
| fleet_age_distribution | [TBD] | ELLIPS equipment records | needs-corporate |
Workshop-Sourced Range: $3-8M/yr Confidence: Medium — concept clear, Pete's Excel models exist, data scattered Key Quote: "At 120,000 hours, we're gonna have engine, two wheel motors, the truck body is going to be worn out — you're over the face value of a new one."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, domain specialist, PM | [TBD] | Fleet lifecycle economics, vendor data mapping |
| Data engineering | Data engineer | [TBD] | ELLIPS + vendor rebuild reports + financial consolidation |
| ML/AI development | Data scientist | [TBD] | Total-cost-of-ownership model, replacement optimizer |
| Application/UX | Frontend dev | [TBD] | Fleet lifecycle dashboard, CAPEX planner |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Moderate — capital planning politics. 20%. |
TLD-P05: Fixed Plant PdM & Failure Analytics¶
TLD-03: Fixed Plant PdM (AG Mills, Kilns, Conveyors)¶
Card Type: B — Structured Corporate Project: PRJ-03
Value Analysis¶
Value Types: Cost avoidance + Throughput gain Value Formula:
(unplanned_mill_shutdowns_per_year × cost_per_shutdown × prevention_rate)
+ (kiln_campaign_life_extension × campaign_value)
+ (lube_system_failures_per_year × cost_per_failure × prevention_rate)
| Variable | Value | Source | Status |
|---|---|---|---|
| AG_mills | 12 + 24 pebble mills | Confirmed | workshop-confirmed |
| mill_shutdown_cost | ~$1M per mill | JR: "$1M per mill shutdown" | workshop-confirmed |
| DCS_breadcrumb_trail | "Signs 3 months before failure" | George Harmon | workshop-confirmed |
| SKF_vibration | Analysts on-site, hardwired sensors | Confirmed | workshop-confirmed |
| Pi_historian_entries | 1.3 billion | Confirmed | workshop-confirmed |
| unplanned_mill_shutdowns_per_year | [TBD] | ELLIPS records | needs-corporate |
| prevention_rate | 20-30% | Conservative H1 | estimated |
| lube_system_failures_per_year | [TBD] | "A lot of problem" with lube systems | needs-corporate |
| cost_per_lube_failure | [TBD] | Repair + downtime | needs-corporate |
Workshop-Sourced Range: $2-5M/yr Confidence: Medium-High — DCS breadcrumb trail = clearest PdM articulation at any site Key Quote: "There were signs of this three months ago. Something had changed. Current increased, then a leaking seal, then vibration work orders, and eventually the part failed."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist, PM | [TBD] | Asset selection, DCS/Pi data audit |
| Data engineering | Data engineer | [TBD] | DCS + SKF + ELLIPS + Pi historian integration |
| ML/AI development | ML engineer, data scientist | [TBD] | Anomaly detection per asset class |
| Application/UX | Frontend dev | [TBD] | Plant health dashboard |
| Infrastructure | Moderate | [TBD] | Filter instrumentation ($125K), lube monitoring |
| Change management | — | [TBD] | Low — George Harmon actively wants this. 15%. |
TLD-42: Cross-Asset Failure Pattern Search¶
Card Type: A — Anchored Corporate Project: PRJ-01
Value Analysis¶
Value Types: Efficiency gain + Cost avoidance Value Formula:
failure_analysis_hours_saved_per_week × weeks_per_year × labor_cost_per_hour
+ proactive_failure_prevention_value (from pattern detection)
| Variable | Value | Source | Status |
|---|---|---|---|
| failure_analysis_time_current | "Hour or two per equipment node" + "hours per week on drawings" | George Harmon | workshop-confirmed |
| drawing_database_size | 60,000 prints | Confirmed | workshop-confirmed |
| ELLIPS_search_quality | "Pretty awkward" | George Harmon | workshop-confirmed |
| failure_analysis_hours_per_week | [TBD] | George Harmon's time allocation | needs-corporate |
| labor_cost_per_hour | [TBD] | Reliability engineer loaded rate | needs-corporate |
| similar_equipment_groups | [TBD] | ELLIPS equipment taxonomy | needs-corporate |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High — pain clearly quantified, well-scoped RAG application Key Quote: "It can take an hour or two looking at the work order history on that particular node. That's without even looking at any of the similar equipment."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | ELLIPS data schema, drawing database format |
| Data engineering | Data engineer | [TBD] | ELLIPS + drawing database ingestion |
| ML/AI development | ML engineer | [TBD] | RAG over work orders + drawings, pattern detection |
| Application/UX | Frontend dev | [TBD] | Natural language search interface |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Low — George's team is the primary user. 15%. |
TLD-41: Deferred Maintenance Risk Quantification¶
Card Type: B — Structured Corporate Project: new
Value Analysis¶
Value Types: Capital optimization + Risk mitigation Value Formula:
(deferred_items × avg_escalation_factor × avg_repair_cost)
+ (budget_misallocation_between_mine_and_plant × correction_value)
| Variable | Value | Source | Status |
|---|---|---|---|
| deferred_maintenance_items | [TBD] | ELLIPS deferred work orders | needs-corporate |
| avg_escalation_factor | [TBD] | Historical: deferred PM to emergency repair cost ratio | needs-corporate |
| avg_repair_cost | [TBD] | ELLIPS cost data | needs-corporate |
| budget_misallocation | [TBD] | Mine vs. plant allocation analysis | needs-corporate |
Workshop-Sourced Range: $1-5M/yr Confidence: Medium — conceptually powerful, needs data to validate Key Quote: "Pay now or pay later. Paying later is almost always more expensive."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, data scientist, PM | [TBD] | Asset criticality framework, cost trajectory analysis |
| Data engineering | Data engineer | [TBD] | ELLIPS + Oracle financial + DCS integration |
| ML/AI development | Data scientist | [TBD] | Cost escalation model, risk scoring |
| Application/UX | Frontend dev | [TBD] | Risk dashboard, budget allocation tool |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | High — changes budget allocation politics. 25%. |
TLD-P06: Drill & Blast Intelligence ★★★¶
TLD-05: Drill & Blast Pattern Optimization¶
Card Type: A — Anchored Corporate Project: new (mining-specific)
Value Analysis¶
Value Types: Direct cost reduction + Throughput gain (downstream) Value Formula:
(annual_explosive_spend × savings_on_soft_holes_%)
+ (grinding_energy_reduction_from_better_fragmentation × energy_cost)
+ (rework_reduction_from_underblasted_zones × rework_cost)
| Variable | Value | Source | Status |
|---|---|---|---|
| drill_holes_per_year | ~15,000 | Jeff Domann: "15,000 drill holes" | workshop-confirmed |
| current_loading_method | Blanket-loaded (same density every hole) | Jeff: "we basically blanket load the patterns" | workshop-confirmed |
| Dyno_auto_density_capability | Confirmed — trucks can auto-load per hole | Jeff: "they have that capability on their trucks now" | workshop-confirmed |
| annual_explosive_spend | [TBD] | Procurement records | needs-corporate |
| savings_on_soft_holes_% | 10-20% | Less explosive needed in soft rock | estimated |
| grinding_energy_savings | [TBD] | Better fragmentation reduces grinding | needs-corporate |
| rework_reduction | [TBD] | Fewer oversize blocks jamming crushers | needs-corporate |
Workshop-Sourced Range: $1-3M/yr Confidence: High ★★★ — both ends of pipeline exist, contractor has capability, data exists per hole Key Quote: "If we could bring that data in, our explosive manufacturer has capabilities on their trucks to know what hole they're pulled up next to. They would automatically know how much to put in."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, PM | [TBD] | Drill data format mapping, Dyno input spec |
| Data engineering | Data engineer | [TBD] | Drill data + Vulcan + Dyno integration |
| ML/AI development | Data scientist | [TBD] | Hardness index from drill metrics, density optimization |
| Application/UX | Frontend dev | [TBD] | Blast plan visualization, per-hole density map |
| Infrastructure | Minimal | [TBD] | Data bridge, no new hardware |
| Change management | — | [TBD] | Low — Dyno already has the truck. Jeff is champion. 15%. |
TLD-53: Drill Consumable Predictive Ordering¶
Card Type: C — Absorbed Corporate Project: new Reason: Low-value add-on to TLD-05. Drill data already captured; consumable forecasting is a minor extension. Value Contribution: $0.2-0.5M/yr — captured in TLD-P06 roll-up. Cost Contribution: One forecasting model within TLD-P06 scope.
TLD-P07: Mine Operations & Dispatch Intelligence ★★★¶
TLD-04: Haul Truck Fleet Dispatching Optimization¶
Card Type: B — Structured Corporate Project: PRJ-07 + PRJ-02 (reframed)
Value Analysis¶
Value Types: Throughput gain + Efficiency gain Value Formula:
fleet_productivity_improvement_% × total_tons_hauled_per_year × value_per_ton
+ dispatch_error_reduction × cost_per_error × errors_per_year
+ blend_compliance_improvement_% × off_blend_cost_per_year
| Variable | Value | Source | Status |
|---|---|---|---|
| truck_count | 14 fleet | Confirmed | workshop-confirmed |
| dispatcher_training_time | "Months to optimize" | Workshop | workshop-confirmed |
| fleet_productivity_improvement_% | 5-15% | Industry benchmark for AI dispatch | estimated |
| total_tons_hauled_per_year | [TBD] | Modular dispatch records | needs-corporate |
| value_per_ton | [TBD] | Pellet margin | needs-corporate |
| dispatch_error_rate | [TBD] | Historical mismatch events | needs-corporate |
Workshop-Sourced Range: $2-6M/yr Confidence: Med-High — Modular data exists, dispatcher pain clearly articulated Key Quote: "What kind of resources are available to them? When they have to sit in the chair and somebody is on vacation..."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, optimization specialist, PM | [TBD] | Dispatch workflow analysis |
| Data engineering | Data engineer | [TBD] | Modular data export, priority ingestion |
| ML/AI development | ML engineer, optimization specialist | [TBD] | Dispatch optimization engine, blend optimizer |
| Application/UX | Frontend dev | [TBD] | Dispatcher decision support interface |
| Infrastructure | Moderate | [TBD] | Real-time optimization engine |
| Change management | — | [TBD] | Moderate — dispatcher workflow change. 20%. |
TLD-50: Real-Time Mine Plan Deviation Alerting¶
Card Type: A — Anchored Corporate Project: new
Value Analysis¶
Value Types: Throughput protection + Knowledge accumulation Value Formula:
plan_deviation_events_per_year × avg_recovery_time_hours × throughput_per_hour × reduction_%
+ undocumented_audible_decisions × knowledge_value_per_decision
| Variable | Value | Source | Status |
|---|---|---|---|
| plan_deviation_discovery_lag | "Next day" currently | Brad Koski, Andrew Mullen | workshop-confirmed |
| shovel_misassignment_example | Drill moved instead of shovel | Brad's specific example | workshop-confirmed |
| plan_deviation_events_per_year | [TBD] | Dispatch vs. plan comparison data | needs-corporate |
| avg_recovery_time_hours | [TBD] | Time to detect + correct deviation | needs-corporate |
| throughput_per_hour | [TBD] | Concentrator throughput | needs-corporate |
Workshop-Sourced Range: $1-3M/yr Confidence: Med-High — multiple leaders articulated the need, data exists in Modular and plan Key Quote: "If we could have something saying real time, hey, you're getting way off plan. This is what I recommend to get back on plan." — Andrew Mullen
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | Plan format parsing, comparison logic |
| Data engineering | Data engineer | [TBD] | Modular dispatch + daily plan ingestion |
| ML/AI development | Data scientist | [TBD] | Comparison engine, deviation classification |
| Application/UX | Frontend dev | [TBD] | Real-time alerting dashboard |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Moderate — accountability implications. 20%. |
TLD-15: Mine Plan & Production Scheduling¶
Card Type: C — Absorbed Corporate Project: PRJ-02 (reframed) Reason: H3 strategic play. JR articulated the cascading vision, but full mine plan scheduling is years out. Plan deviation alerting (TLD-50) is the achievable H2 entry point. Value Contribution: $3-8M/yr — captured as the H3 ceiling for TLD-P07. Cost Contribution: Significant — full optimization engine, multi-system integration.
TLD-26: Operator Performance & Payload Analytics¶
Card Type: C — Absorbed Corporate Project: new Reason: Subset of dispatch intelligence. Automated scorecards from existing Modular data. Value Contribution: $1-3M/yr (reduced overloading damage) — captured in TLD-P07 roll-up. Cost Contribution: One analytics module within TLD-P07 scope.
TLD-P08: Mine-to-Dock Logistics Optimization ★★★¶
TLD-16: Vessel/Shipping Schedule & Rail Coordination¶
Card Type: A — Anchored Corporate Project: PRJ-07 (reframed)
Value Analysis¶
Value Types: Efficiency gain + Cost avoidance Value Formula:
(daily_replanning_hours × labor_cost_per_hour × 365)
+ (wasted_train_crew_deployments_per_year × crew_cost_per_deployment)
+ (dock_utilization_improvement_% × annual_throughput × margin_per_ton)
+ (demurrage_cost_reduction)
| Variable | Value | Source | Status |
|---|---|---|---|
| daily_replanning_hours | 3-4 hours/day | Kevin: "3-4 hours every day replanning" | workshop-confirmed |
| BCS_data_available | Years of shipping history | "Endless amount of vessel histories" | workshop-confirmed |
| dock_age | 130 years, single-source failure | Confirmed | workshop-confirmed |
| vessel_contractors | 3 contractors, 4-6 vessels | Confirmed | workshop-confirmed |
| schedule_rolling_window | 30-day rolling, daily changes | Kevin confirmed | workshop-confirmed |
| labor_cost_per_hour | [TBD] | Loaded scheduling/logistics rate | needs-corporate |
| wasted_crew_deployments_per_year | [TBD] | When vessels don't show | needs-corporate |
| crew_cost_per_deployment | [TBD] | Train crew loaded cost | needs-corporate |
| dock_utilization_current | [TBD] | BCS data | needs-corporate |
| annual_throughput | ~7.7M tons pellets | Production target | workshop-confirmed |
| demurrage_cost | [TBD] | Vessel waiting costs | needs-corporate |
Workshop-Sourced Range: $2-5M/yr Confidence: High — massive BCS data, clear daily pain articulated unprompted Key Quotes: "That's probably our biggest business challenge this year." "Every day, having to keep replanning somebody's train crews based on the vessel schedule."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | BCS data audit, scheduling workflow mapping |
| Data engineering | Data engineer | [TBD] | BCS + SharePoint + weather + dock status integration |
| ML/AI development | ML engineer, data scientist | [TBD] | Schedule optimizer (L1), disruption predictor (L2) |
| Application/UX | Frontend dev | [TBD] | Daily schedule dashboard, crew call board |
| Infrastructure | Minimal | [TBD] | BCS already has the data |
| Change management | — | [TBD] | Low — Kevin is the user and wants this. 15%. |
TLD-37: Railroad Asset Maintenance Analytics¶
Card Type: C — Absorbed Corporate Project: PRJ-03 Reason: Team self-assessed "data not there yet." Geo car + X-ray car data exist but ELLIPS railroad data is limited. Value Contribution: $0.3-1M/yr — captured in TLD-P08 roll-up. Cost Contribution: Phase 3 add-on after scheduling optimizer.
TLD-P09: Ops-Maintenance Data Integration¶
TLD-01: Mining Ops-Maintenance Data Integration¶
Card Type: A — Anchored Corporate Project: PRJ-01
Value Analysis¶
Value Types: Throughput gain + Efficiency gain Value Formula:
misattributed_delay_hours_per_month × production_value_per_hour × attribution_correction_rate
+ manual_data_entry_hours_per_week × labor_cost_per_hour × 52
+ dispatch_data_correction_hours_per_week × labor_cost_per_hour × 52
| Variable | Value | Source | Status |
|---|---|---|---|
| CMMS | ELLIPS (3rd different CMMS across CLF) | Confirmed | workshop-confirmed |
| parallel_systems | ELLIPS, DCS, drawing DB (60K), relay system, Business Objects, Power BI, Oracle | George Harmon: "none of them talking to each other" | workshop-confirmed |
| manual_hours_entry | 4 hours every Monday | George Beelon confirmed | workshop-confirmed |
| PM_rate | 70/30 PM/reactive (better than steel sites) | Maintenance team confirmed | workshop-confirmed |
| corporate_validation | Andrew Mullen: "Doesn't matter what CMMS — it's not getting done" | 3/3 sites | workshop-confirmed |
| production_value_per_hour | [TBD] | Concentrator throughput x pellet margin | needs-corporate |
| labor_cost_per_hour | [TBD] | Loaded maintenance tech rate | needs-corporate |
Workshop-Sourced Range: $2-5M/yr Confidence: High — pattern validated at 3/3 sites, now corporate-confirmed by Andrew Mullen Key Quote: "Doesn't matter what CMMS it is... it's not getting done because it's just too cumbersome." — Andrew Mullen
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | System integration mapping |
| Data engineering | Data engineer (senior) | [TBD] | ELLIPS + DCS + Modular + Business Objects integration |
| ML/AI development | ML engineer | [TBD] | Semantic matching, delay attribution |
| Application/UX | Frontend dev | [TBD] | Unified ops-maint dashboard |
| Infrastructure | Moderate | [TBD] | Data layer (Microsoft Fabric likely) |
| Change management | — | [TBD] | Moderate — ops + maint alignment. 20%. |
TLD-45: Modular Dispatch to ELLIPS Automated Integration¶
Card Type: A — Anchored Corporate Project: PRJ-01
Value Analysis¶
Value Types: Efficiency gain + Data quality Value Formula:
manual_entry_hours_per_week × weeks_per_year × labor_cost_per_hour
+ PM_scheduling_accuracy_improvement × maintenance_spend_affected
+ ELLIPS_prediction_error_cost (from flawed hour averaging)
| Variable | Value | Source | Status |
|---|---|---|---|
| manual_entry_hours | 4 hours/week | "About 4 hours entering those hours" | workshop-confirmed |
| ELLIPS_prediction_flaw | Pushes PMs out when machine was down, not running less | Confirmed | workshop-confirmed |
| Modular_data_available | Hours, keys on/off, loaded/unloaded, GPS, fuel | Confirmed | workshop-confirmed |
| labor_cost_per_hour | [TBD] | Loaded scheduler rate | needs-corporate |
| PM_scheduling_accuracy_impact | [TBD] | Cost of mis-timed PMs | needs-corporate |
Workshop-Sourced Range: $1-3M/yr Confidence: High — both systems have good data, purely an integration gap Key Quote: "The information's there. They just don't know how to get it into ELLIPS."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | Data mapping: Modular fields to ELLIPS meters |
| Data engineering | Data engineer | [TBD] | API/export integration, validation rules |
| ML/AI development | Minimal | [TBD] | Data validation, anomaly flagging |
| Application/UX | Minimal | [TBD] | Status monitoring dashboard |
| Infrastructure | Minimal | [TBD] | Data pipeline |
| Change management | — | [TBD] | Low — eliminates manual work. 15%. |
TLD-49: Dispatch Status Auto-Correction¶
Card Type: A — Anchored Corporate Project: PRJ-01
Value Analysis¶
Value Types: Data quality + Efficiency gain Value Formula:
correction_hours_per_week × weeks_per_year × labor_cost_per_hour
+ downstream_data_quality_improvement_value
| Variable | Value | Source | Status |
|---|---|---|---|
| correction_method | Molly manually scans 12-hour shifts | Kevin confirmed | workshop-confirmed |
| pattern_complexity | "Easy thing to spot, but tedious" | Kevin | workshop-confirmed |
| correction_hours_per_week | [TBD] | Molly's time allocation | needs-corporate |
| labor_cost_per_hour | [TBD] | Dispatch admin loaded rate | needs-corporate |
Workshop-Sourced Range: $0.3-1M/yr Confidence: High — straightforward anomaly detection, training data exists from historical corrections Key Quote: "I can see exactly where they missed this button. It's a really actually easy thing to spot, but it's just tedious."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist | [TBD] | Button-press pattern analysis |
| Data engineering | Data engineer | [TBD] | Modular dispatch data access |
| ML/AI development | ML engineer | [TBD] | Anomaly detection / rules-based correction |
| Application/UX | Frontend dev | [TBD] | Correction review interface for Molly |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Low — Molly wants this. 10%. |
TLD-P10: HPGR Knowledge Base + PdM Pilot ★★★ LEAD PILOT¶
TLD-38: HPGR Knowledge Base + PdM Pilot¶
Card Type: A — Anchored Corporate Project: PRJ-06 + PRJ-03
Value Analysis¶
Value Types: Efficiency gain + Cost avoidance + Knowledge preservation Value Formula:
(troubleshooting_time_saved_per_event × events_per_year × labor_cost_per_hour)
+ (HPGR_unplanned_downtime_hours × prevention_rate × throughput_value_per_hour)
+ (knowledge_transfer_risk_avoidance_for_new_equipment)
| Variable | Value | Source | Status |
|---|---|---|---|
| unread_manuals | 10+ manuals, 1,200+ pages | "Nobody here has read" | workshop-confirmed |
| troubleshooting_time_current | "Always takes a couple days" | Adam Bingham | workshop-confirmed |
| sensor_coverage | "Covered in sensors" | Confirmed | workshop-confirmed |
| champion_status | Adam Bingham already using Copilot | Confirmed — proof-of-concept in production | workshop-confirmed |
| team_nominated | Yes — maintenance team consensus | Day 2 Plant Maintenance | workshop-confirmed |
| troubleshooting_events_per_year | [TBD] | ELLIPS work orders since April 2023 | needs-corporate |
| labor_cost_per_hour | [TBD] | Maintenance tech loaded rate | needs-corporate |
| HPGR_unplanned_downtime_hours | [TBD] | ELLIPS downtime records | needs-corporate |
| throughput_value_per_hour | [TBD] | Concentrator throughput x pellet margin | needs-corporate |
| prevention_rate | 20-30% | Conservative for new equipment with limited history | estimated |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High ★★★ — team-nominated, documentation digital, sensors confirmed, champion active Key Quotes: "There's literally hundreds of drawings... 1,200 pages of information that no one here has read." "It's covered in sensors, right?"
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | Manual inventory, sensor data audit |
| Data engineering | Data engineer | [TBD] | SharePoint + OEM system + DCS integration |
| ML/AI development | ML engineer | [TBD] | Knowledge base (RAG), anomaly detection models |
| Application/UX | Frontend dev | [TBD] | Copilot-based troubleshooting interface |
| Infrastructure | Minimal | [TBD] | Copilot already available, manuals digital |
| Change management | — | [TBD] | Very low — Adam Bingham already building prototypes. 10%. |
TLD-33: HPGR Feed Rate Root Cause Analysis¶
Card Type: B — Structured Corporate Project: PRJ-03
Value Analysis¶
Value Types: Throughput gain Value Formula:
sustained_feed_rate_improvement × concentrator_throughput_per_unit_rate × pellet_value_per_ton
| Variable | Value | Source | Status |
|---|---|---|---|
| feed_rate_drop_period | Nov 2025, struggled 7-8 months | Sean Halston | workshop-confirmed |
| smoking_gun_status | Not found despite extensive analysis | "Can't say I found the smoking gun" | workshop-confirmed |
| feed_rate_impact | [TBD] | DCS historical data | needs-corporate |
| throughput_value | [TBD] | Feed rate to tons through concentrator | needs-corporate |
Workshop-Sourced Range: $1-3M/yr Confidence: Medium — data exists, but no guarantee ML finds what humans couldn't Key Quote: "With all the data we have available, I can't say I found the smoking gun for it."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist | [TBD] | Hypothesis mapping with engineering team |
| Data engineering | Data engineer | [TBD] | DCS data extraction April 2023 to present |
| ML/AI development | Data scientist (senior) | [TBD] | Multi-variable correlation analysis |
| Application/UX | Minimal | [TBD] | Analysis report |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Low — investigation project. 10%. |
TLD-P11: Maintenance Workflow & Inventory Intelligence¶
TLD-35: ELLIPS Inventory Master Data Cleanup¶
Card Type: A — Anchored Corporate Project: PRJ-06
Value Analysis¶
Value Types: Efficiency gain + Inventory optimization Value Formula:
(part_matching_time_per_box × boxes_per_day × labor_cost_per_hour × 365)
+ (duplicate_inventory_value × carrying_cost_%)
+ (new_duplicate_prevention_value)
| Variable | Value | Source | Status |
|---|---|---|---|
| part_matching_time | 5 minutes to 2 hours per box | Warehouse team | workshop-confirmed |
| ELLIPS_search_quality | "Terrible" — descriptions with misplaced commas/semicolons | Warehouse team | workshop-confirmed |
| duplicate_entries | "12 or 15 or 20 things that are the same thing, but spelled differently" | Warehouse team | workshop-confirmed |
| boxes_per_day | [TBD] | Warehouse receiving volume | needs-corporate |
| labor_cost_per_hour | [TBD] | Warehouse staff loaded rate | needs-corporate |
| total_inventory_value | [TBD] | ELLIPS inventory data | needs-corporate |
| carrying_cost_% | 25% | Industry standard | estimated |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High — identical pattern validated at Middletown (MDT-31), proven recipe Key Quote: "Five minutes to two hours per box or per item that came in trying to find it in the system."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, PM | [TBD] | ELLIPS data export, schema analysis |
| Data engineering | Data engineer | [TBD] | ELLIPS data extraction, NLP pipeline |
| ML/AI development | Data scientist | [TBD] | Semantic dedup, description normalization, search |
| Application/UX | Frontend dev | [TBD] | Natural language search interface |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Low — ops validated: "huge win." 15%. |
TLD-48: OEM Parts Catalog & PM Procedure Auto-Import¶
Card Type: A — Anchored Corporate Project: PRJ-06
Value Analysis¶
Value Types: Efficiency gain + Data quality Value Formula:
new_equipment_onboarding_time_saved × equipment_purchases_per_year × labor_cost_per_hour
+ prevented_duplicate_stock_codes × carrying_cost_per_duplicate
+ faster_maintenance_readiness_value
| Variable | Value | Source | Status |
|---|---|---|---|
| parts_per_new_truck | ~5,000 | Pete: "5,000 different part numbers on that truck" | workshop-confirmed |
| current_process | Manual stock code creation + manual PM procedure entry | Chase Lincoln confirmed | workshop-confirmed |
| stock_code_permissions | Only 2-3 people can create | Confirmed | workshop-confirmed |
| equipment_purchases_per_year | [TBD] | Capital planning records | needs-corporate |
| labor_cost_per_hour | [TBD] | Planner loaded rate | needs-corporate |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High — structured data matching is mature NLP, OEM catalogs are digital Key Quote: "Why do I have to go in and create Cliff's own stock code for each one of those parts?"
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | OEM catalog format mapping |
| Data engineering | Data engineer | [TBD] | OEM catalog ingestion, ELLIPS import |
| ML/AI development | Data scientist | [TBD] | Cross-reference matching, PM procedure parsing |
| Application/UX | Frontend dev | [TBD] | Import review/approval interface |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Moderate — data governance approval needed. 20%. |
TLD-12: Maintenance Workflow Digitization (Copilot)¶
Card Type: B — Structured Corporate Project: PRJ-06
Value Analysis¶
Value Types: Efficiency gain + Knowledge capture Value Formula:
diagnosis_time_saved_per_repair × repairs_per_month × labor_cost_per_hour
+ documentation_improvement_value
| Variable | Value | Source | Status |
|---|---|---|---|
| Adam_Bingham_proof_of_concept | Already using Copilot for troubleshooting + translation | Confirmed | workshop-confirmed |
| ELLIPS_robust_but_cumbersome | Gary: "too cumbersome for people to come back and manually work" | Confirmed | workshop-confirmed |
| repairs_per_month | [TBD] | ELLIPS work order volume | needs-corporate |
| labor_cost_per_hour | [TBD] | Maintenance tech loaded rate | needs-corporate |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High — Adam Bingham = strongest grassroots AI adoption at any site
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, UX researcher, PM | [TBD] | Field shadowing with Adam Bingham |
| Data engineering | Data engineer | [TBD] | ELLIPS + drawing DB + Pi historian ingestion |
| ML/AI development | ML engineer | [TBD] | Voice-to-structured (LLM/STT), RAG over repair history |
| Application/UX | Frontend dev, mobile dev | [TBD] | Voice-first mobile interface |
| Infrastructure | Moderate | [TBD] | STT/LLM inference, connectivity in plant |
| Change management | — | [TBD] | Moderate — trust + UX. 20%. Union. |
TLD-13: Procurement Automation (Parts & Consumables)¶
Card Type: B — Structured Corporate Project: PRJ-06
Value Analysis¶
Value Types: Cost avoidance + Efficiency gain Value Formula:
(parts_delay_frequency × downtime_cost_per_delay × reduction_%)
+ (inventory_right_sizing_savings × total_consumable_spend)
| Variable | Value | Source | Status |
|---|---|---|---|
| parts_delay_frequency | "Weekly" | Gary confirmed: "do you face delays? Weekly." | workshop-confirmed |
| min_max_system_flaw | Doesn't understand set sizes (e.g., 10 injectors per engine) | Pete Austin | workshop-confirmed |
| parts_go_inactive_after_1_year | ELLIPS deactivates unused stock codes | Confirmed | workshop-confirmed |
| total_consumable_spend | [TBD] | ELLIPS + procurement records | needs-corporate |
| downtime_cost_per_delay | [TBD] | Equipment downtime from parts wait | needs-corporate |
Workshop-Sourced Range: $1-3M/yr Confidence: High — validated at MDT as self-funding starter project
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | Procurement workflow mapping |
| Data engineering | Data engineer | [TBD] | ELLIPS inventory + Oracle procurement |
| ML/AI development | Data scientist | [TBD] | Lead time prediction, min/max optimization, set-size awareness |
| Application/UX | Frontend dev | [TBD] | Procurement dashboard, auto-reorder |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Moderate — procurement policy changes. 20%. |
TLD-30: Parts Warehouse Digitization (Barcode/Scanner)¶
Card Type: C — Absorbed Corporate Project: PRJ-06 Reason: Infrastructure enabler for TLD-35 and TLD-13. Straightforward technology deployment. Value Contribution: $0.2-0.5M/yr — captured in TLD-P11 roll-up. Cost Contribution: Hardware procurement + ELLIPS configuration.
TLD-P12: Mining Knowledge Capture & Virtual SME¶
TLD-14: Mining Knowledge Capture / Virtual SME¶
Card Type: A — Anchored Corporate Project: Virtual SME (cross-site)
Value Analysis¶
Value Types: Knowledge preservation + Efficiency gain + Risk mitigation Value Formula:
(troubleshooting_time_saved × incidents_per_year × labor_cost_per_hour)
+ (retirement_knowledge_risk × affected_roles × replacement_cost)
+ (training_acceleration × new_hires_per_year × training_cost_per_hire)
| Variable | Value | Source | Status |
|---|---|---|---|
| Pi_historian_entries | 1.3 billion | Confirmed | workshop-confirmed |
| unread_equipment_manuals | "Hundreds" + 1,200+ pages for HPGR | Confirmed | workshop-confirmed |
| experience_drain | "Most experienced shift manager — five minutes" | Process engineering | workshop-confirmed |
| decision_tree_charts | "80% right, 20% wrong — people learn the chart, not the job" | Confirmed | workshop-confirmed |
| Adam_Bingham_prototype | Already using Copilot for knowledge base | Confirmed | workshop-confirmed |
| incidents_per_year | [TBD] | ELLIPS + safety records | needs-corporate |
| labor_cost_per_hour | [TBD] | Loaded tech rate | needs-corporate |
| new_hires_per_year | [TBD] | HR records | needs-corporate |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High ★★★ — strongest knowledge capture case at any site, grassroots champion already experimenting Key Quote: "We have hundreds of equipment manuals. I don't think anyone here has ever cracked one open."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | Knowledge domain mapping |
| Data engineering | Data engineer | [TBD] | Manual ingestion, Pi historian integration |
| ML/AI development | ML engineer | [TBD] | RAG knowledge base, search optimization |
| Application/UX | Frontend dev | [TBD] | Copilot-based search interface |
| Infrastructure | Minimal | [TBD] | Copilot already available |
| Change management | — | [TBD] | Low — Adam Bingham already adopted. 15%. |
TLD-52: Labor/BLA Contract Knowledge Assistant¶
Card Type: A — Anchored Corporate Project: Virtual SME
Value Analysis¶
Value Types: Efficiency gain + Risk mitigation Value Formula:
supervisor_contract_questions_per_week × time_per_question × labor_cost × 52
+ grievance_reduction × cost_per_grievance
| Variable | Value | Source | Status |
|---|---|---|---|
| availability_gap | No one on-site 24/7 for contract questions | Lynn Casco confirmed | workshop-confirmed |
| new_supervisor_vulnerability | "Guys will try and fool them" | Brad Koski | workshop-confirmed |
| supervisor_contract_questions_per_week | [TBD] | Lynn Casco estimate | needs-corporate |
| grievance_rate | [TBD] | HR records | needs-corporate |
| cost_per_grievance | [TBD] | HR/legal | needs-corporate |
Workshop-Sourced Range: $0.2-0.5M/yr Confidence: High — RAG over single document corpus is a proven pattern, quick win Key Quote: "It would be amazing if they could ask somewhere, a ChatGPT type situation, what do we do now?" — Lynn Casco
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | PM | [TBD] | Contract scope review, legal approval |
| Data engineering | Data engineer | [TBD] | Contract document ingestion |
| ML/AI development | ML engineer | [TBD] | RAG chatbot, contract clause referencing |
| Application/UX | Frontend dev | [TBD] | Mobile-friendly chatbot interface |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Moderate — legal/HR/union approval. 20%. |
TLD-51: Shift Handover & Ops Knowledge Base¶
Card Type: B — Structured Corporate Project: PRJ-01 + Virtual SME
Value Analysis¶
Value Types: Knowledge accumulation + Efficiency gain Value Formula:
repeat_mistake_cost × repeat_incidents_per_year × reduction_%
+ shift_transition_time_saved × shifts_per_year × labor_cost
| Variable | Value | Source | Status |
|---|---|---|---|
| shift_email_quality | "Varies wildly by supervisor" | Brad Koski, Dan Kernan | workshop-confirmed |
| knowledge_loss | "A lot of it just lives in our memories" | Dan Kernan | workshop-confirmed |
| repeat_incidents_per_year | [TBD] | Ops records | needs-corporate |
| labor_cost | [TBD] | Loaded supervisor rate | needs-corporate |
Workshop-Sourced Range: $0.5-2M/yr Confidence: High — data sources exist (shift emails + dispatch PDFs), proven NLP pattern
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | Shift email template analysis |
| Data engineering | Data engineer | [TBD] | Email + dispatch PDF + Modular integration |
| ML/AI development | ML engineer | [TBD] | Auto-summarization, prompted documentation, RAG |
| Application/UX | Frontend dev | [TBD] | Shift knowledge dashboard |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Moderate — supervisor documentation habits. 20%. |
TLD-43: Maintenance Training Content Generation¶
Card Type: C — Absorbed Corporate Project: Virtual SME Reason: Extension of TLD-14 knowledge base. Visual training content generated from the same manual corpus. Value Contribution: $0.3-1M/yr — captured in TLD-P12 roll-up. Cost Contribution: One content generation pipeline within TLD-P12 scope.
TLD-27: Environmental Compliance Knowledge System¶
Card Type: C — Absorbed Corporate Project: new Reason: Specific instance of knowledge capture theme. Environmental compliance knowledge from "2-3 heads" into system. Value Contribution: $0.3-1M/yr — captured in TLD-P12 roll-up. Cost Contribution: One knowledge domain within TLD-P12 scope.
TLD-P13: Maintenance Planning & Scheduling¶
TLD-39: Major Repair Schedule Optimization¶
Card Type: A — Anchored Corporate Project: new
Value Analysis¶
Value Types: Efficiency gain + Throughput gain Value Formula:
supervisor_scheduling_hours_per_week × weeks_per_year × labor_cost_per_hour
+ (repair_coordination_improvement × mill_shutdowns_per_year × cost_per_shutdown × reduction_%)
| Variable | Value | Source | Status |
|---|---|---|---|
| mill_shutdown_cost | ~$1M per mill | JR confirmed | workshop-confirmed |
| current_process | Manual ELLIPS to Excel to MS Project, daily updates | Gary, Steve | workshop-confirmed |
| seasonal_constraint | Mid-March through June — no large parts deliverable | Confirmed | workshop-confirmed |
| supervisor_scheduling_hours_per_week | [TBD] | Senior supervisor time allocation | needs-corporate |
| mill_shutdowns_per_year | [TBD] | Maintenance records | needs-corporate |
| labor_cost_per_hour | [TBD] | Senior supervisor loaded rate | needs-corporate |
Workshop-Sourced Range: $1-3M/yr Confidence: Med-High — "$1M per mill shutdown" is a clear anchor Key Quote: "Our senior supervisors are updating the line repairs every day, trying to get an end date. If you could make that quick, so they're analyzing data not inputting data — that would be a huge win."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Solution architect, PM | [TBD] | Repair scheduling workflow mapping |
| Data engineering | Data engineer | [TBD] | ELLIPS to MS Project automation |
| ML/AI development | Data scientist | [TBD] | Critical chain optimizer, resource conflict detection |
| Application/UX | Frontend dev | [TBD] | Auto-updating schedule dashboard |
| Infrastructure | Minimal | [TBD] | ELLIPS + MS Project integration |
| Change management | — | [TBD] | Low — supervisors actively requesting this. 15%. |
TLD-40: Maintenance Resource & Workforce Scheduling¶
Card Type: C — Absorbed Corporate Project: new Reason: Subset of TLD-39 scheduling scope. Daily crew assignment is an extension of major repair scheduling. Value Contribution: $0.5-2M/yr — captured in TLD-P13 roll-up. Cost Contribution: One scheduling module within TLD-P13 scope.
TLD-36: Maintenance Parts & Budget Forecasting¶
Card Type: B — Structured Corporate Project: new
Value Analysis¶
Value Types: Budget accuracy + Cost avoidance Value Formula:
budget_variance_% × total_maintenance_spend × variance_cost_factor
+ emergency_procurement_events × premium_per_event
| Variable | Value | Source | Status |
|---|---|---|---|
| current_method | Straight-line averages, 65+ item categories | JR confirmed | workshop-confirmed |
| untracked_small_items | $300K/yr "Walmart effect" | JR confirmed | workshop-confirmed |
| total_maintenance_spend | [TBD] | Oracle financial records | needs-corporate |
| budget_variance_% | [TBD] | Historical budget vs. actual | needs-corporate |
| emergency_procurement_premium | [TBD] | Premium for rush orders | needs-corporate |
Workshop-Sourced Range: $0.5-2M/yr Confidence: Med-High — clear pain, data exists in ELLIPS + production model Key Quote: "Straight-line averages are the tough ones."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | Data scientist, PM | [TBD] | Correlation analysis design |
| Data engineering | Data engineer | [TBD] | ELLIPS + production model + Oracle integration |
| ML/AI development | Data scientist | [TBD] | Production-correlated spend model, what-if scenarios |
| Application/UX | Frontend dev | [TBD] | Budget forecasting dashboard |
| Infrastructure | Minimal | [TBD] | — |
| Change management | — | [TBD] | Low — budget owners want better forecasts. 15%. |
TLD-P14: Workplace Safety & Inspection Digitization¶
TLD-24: Workplace & Equipment Inspection Digitization¶
Card Type: A — Anchored Corporate Project: new (MDT-P07 parallel)
Value Analysis¶
Value Types: Efficiency gain + Compliance improvement Value Formula:
(cards_per_shift × shifts_per_day × days_per_year × processing_time_per_card × labor_cost_per_hour)
+ (corrective_action_completion_improvement × incident_prevention_value)
+ (equipment_inspection_time_saved × inspections_per_day × labor_cost_per_hour)
| Variable | Value | Source | Status |
|---|---|---|---|
| cards_per_shift | ~50 | "He has fifty cards in his hand" | workshop-confirmed |
| shifts_per_day | 3 (24/7 operations) | Confirmed | workshop-confirmed |
| equipment_inspection_time | Up to 2 hours on paper | Confirmed | workshop-confirmed |
| corrective_action_tracking | None — no reminders, no follow-up | Confirmed | workshop-confirmed |
| processing_time_per_card | [TBD] | Supervisor data entry time | needs-corporate |
| labor_cost_per_hour | [TBD] | Supervisor loaded rate | needs-corporate |
| incident_prevention_value | [TBD] | MSHA fine history + incident cost | needs-corporate |
Workshop-Sourced Range: $0.3-1M/yr Confidence: High — strong group energy, proven mobile capture technology Key Quote: "Directions to be as simple as taking your smartphone and capturing a video or some pictures. All that information would flow freely."
Cost Analysis¶
| Component | Vooban Team | IE | Notes |
|---|---|---|---|
| Discovery & design | UX researcher, PM | [TBD] | Take-5 workflow mapping |
| Data engineering | Data engineer | [TBD] | Voice/photo to structured data pipeline |
| ML/AI development | ML engineer | [TBD] | Voice-to-structured extraction, corrective action tracking |
| Application/UX | Mobile dev | [TBD] | Smartphone capture app |
| Infrastructure | Moderate | [TBD] | Connectivity in pit/plant needed |
| Change management | — | [TBD] | Low — workers want simpler process. 15%. |
TLD-20: Safety Analytics¶
Card Type: C — Absorbed Corporate Project: new Reason: Depends on TLD-24 digital capture as data foundation. Analytics layer on top of digitized safety data. Value Contribution: $0.5-2M/yr — captured in TLD-P14 roll-up. Cost Contribution: Analytics module within TLD-P14 scope.
TLD-25: Mine Production Reporting Automation¶
Card Type: C — Absorbed Corporate Project: new Reason: Same data capture + automation pattern as TLD-24. Paper to digital reporting. Value Contribution: $0.2-0.5M/yr — captured in TLD-P14 roll-up. Cost Contribution: One reporting pipeline within TLD-P14 scope.
TLD-P15: Environmental, Utilities & Geotechnical¶
TLD-18: Environmental Compliance Analytics¶
Card Type: C — Absorbed Corporate Project: new Reason: Seed status — limited field evidence. Environmental knowledge capture is in TLD-P12 (Virtual SME). Analytics layer needs deeper scoping. Value Contribution: $0.5-2M/yr estimated — captured in TLD-P15 roll-up. Cost Contribution: One monitoring/prediction module.
TLD-28: Utilities/Energy Consumption Forecasting¶
Card Type: C — Absorbed Corporate Project: new Reason: Seed — clear pain but lower priority than production. Power contract complexity mentioned but not scoped. Value Contribution: $0.5-2M/yr estimated — captured in TLD-P15 roll-up. Cost Contribution: One forecasting model.
TLD-17: Haul Road & Pit Slope Monitoring¶
Card Type: C — Absorbed Corporate Project: new Reason: Seed — depends on geotechnical monitoring infrastructure that hasn't been scoped. Value Contribution: $0.5-2M/yr estimated — captured in TLD-P15 roll-up. Cost Contribution: Depends on sensor infrastructure assessment.
TLD-P16: HR & Administrative Operations¶
TLD-29: HR/Workforce Overtime Forecasting¶
Card Type: C — Absorbed Corporate Project: new Reason: Low strategic priority. Clear pain but minimal production impact. Value Contribution: $0.2-0.5M/yr — captured in TLD-P16 roll-up. Cost Contribution: One forecasting model.
TLD-44: Employee Onboarding Automation¶
Card Type: C — Absorbed Corporate Project: new Reason: IT process automation. Low strategic value relative to operational initiatives. Value Contribution: $0.1-0.5M/yr — captured in TLD-P16 roll-up. Cost Contribution: ServiceNow integration.
Project Roll-Ups¶
| Project | Initiatives | Anchored (A) | Structured (B) | Absorbed (C) | Workshop Range | Confidence |
|---|---|---|---|---|---|---|
| TLD-P01 Concentrator Feed-Forward & Ore Intelligence ★★★ | TLD-21, TLD-06, TLD-31, TLD-23 | 3 | 1 | 0 | $8-16M/yr | High |
| TLD-P02 Concentrator Ops & Recovery | TLD-07, TLD-08, TLD-22, TLD-32, TLD-11 | 3 | 1 | 1 | $7-16M/yr | Med-High |
| TLD-P03 Pellet Plant Quality & Control | TLD-34, TLD-09, TLD-10 | 1 | 1 | 1 | $3-9M/yr | Med-High |
| TLD-P04 Mining Fleet PdM & Lifecycle ★★★ | TLD-19, TLD-02, TLD-46, TLD-47 | 2 | 2 | 0 | $9-24M/yr | High |
| TLD-P05 Fixed Plant PdM & Failure Analytics | TLD-03, TLD-42, TLD-41 | 1 | 2 | 0 | $4-12M/yr | Med-High |
| TLD-P06 Drill & Blast Intelligence ★★★ | TLD-05, TLD-53 | 1 | 0 | 1 | $1-4M/yr | High |
| TLD-P07 Mine Ops & Dispatch Intelligence ★★★ | TLD-04, TLD-50, TLD-15, TLD-26 | 1 | 1 | 2 | $6-20M/yr | Med-High |
| TLD-P08 Mine-to-Dock Logistics ★★★ | TLD-16, TLD-37 | 1 | 0 | 1 | $2-6M/yr | High |
| TLD-P09 Ops-Maint Data Integration | TLD-01, TLD-45, TLD-49 | 3 | 0 | 0 | $4-9M/yr | High |
| TLD-P10 HPGR Pilot ★★★ | TLD-38, TLD-33 | 1 | 1 | 0 | $2-5M/yr | High |
| TLD-P11 Maint Workflow & Inventory | TLD-35, TLD-48, TLD-12, TLD-13, TLD-30 | 2 | 2 | 1 | $3-10M/yr | High |
| TLD-P12 Knowledge Capture & Virtual SME | TLD-14, TLD-52, TLD-51, TLD-43, TLD-27 | 2 | 1 | 2 | $2-7M/yr | High |
| TLD-P13 Maintenance Planning & Scheduling | TLD-39, TLD-40, TLD-36 | 1 | 1 | 1 | $2-7M/yr | Med-High |
| TLD-P14 Safety & Inspection Digitization | TLD-24, TLD-20, TLD-25 | 1 | 0 | 2 | $1-4M/yr | High |
| TLD-P15 Environmental, Utilities & Geotechnical | TLD-18, TLD-28, TLD-17 | 0 | 0 | 3 | $1-4M/yr | Low-Med |
| TLD-P16 HR & Administrative | TLD-29, TLD-44 | 0 | 0 | 2 | $0.3-1M/yr | Medium |
| TOTAL | 53 | 23 | 13 | 17 | $50-153M/yr |
Card type distribution: 23 Anchored (43%), 13 Structured (25%), 17 Absorbed (32%). The 36 cards with formulas (A+B) cover the bulk of the value — the 17 Absorbed initiatives contribute within parent projects.
Corporate Inquiry Table — Tilden Mine¶
Purpose: All variables tagged
needs-corporatein one table. Send to IE for Cleveland-Cliffs data request.
Production & Revenue¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 1 | pellet_value_per_ton | TLD-21, TLD-06, TLD-08, TLD-23, TLD-31, TLD-32, TLD-46, TLD-50 | What is the average pellet market value per ton? (or margin per ton by customer if available) | Critical — used across 8+ cards |
| 2 | annual_pellet_tons (actual) | TLD-08, TLD-09, TLD-23 | What was Tilden's actual pellet production in the last 12 months? (vs. the ~7.7M target) | Critical |
| 3 | throughput_value_per_hour (concentrator) | TLD-21, TLD-22, TLD-38, TLD-01 | What is the concentrator throughput value per operating hour? (tons/hour x pellet margin) | Critical |
Reagent & Chemical Spend¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 4 | reagent_spend_breakdown_by_type | TLD-23 | What is the breakdown of the ~$50M/yr reagent spend by type? (amine, corn starch, depressants, pH modifiers, etc.) | High |
| 5 | reagent_dosing_data_availability | TLD-21, TLD-23 | Is reagent dosing logged in DCS? At what frequency? How far back? | High |
| 6 | per_section_reagent_variance | TLD-31 | Is there data on reagent consumption variation between concentrator sections (2-3, 4-6, 7-9, 10-12)? | Medium |
Concentrator & Pellet Plant Operations¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 7 | current_throughput_per_mill | TLD-07 | What is the average throughput (tons/hour) per AG mill? What is the variability range? | High |
| 8 | energy_cost_per_kWh | TLD-07, TLD-11 | What is Tilden's effective energy cost per kWh? (including power contract structure) | High |
| 9 | filter_degradation_events_per_year | TLD-22 | How many filter degradation events per year? Average detection time? Average throughput impact? | High |
| 10 | six_bearing_events_per_year | TLD-32 | How frequently do six-bearing recirculation events occur? (per shift, per day, per week?) | Medium |
| 11 | off_spec_pellet_rate | TLD-09 | What percentage of pellets fail quality specs? (compressive strength, chemistry) | Medium |
| 12 | annual_energy_cost_pellet_plant | TLD-10 | What is the annual energy cost for the pellet plant? (natural gas, electricity breakdown) | Medium |
Fleet & Mobile Equipment¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 13 | unplanned_failures_per_year (fleet) | TLD-02 | How many unplanned truck/shovel breakdowns occurred in the last 12 months? Average downtime per event? | High |
| 14 | avg_repair_cost (fleet) | TLD-02 | What is the average repair cost per unplanned fleet event? (by equipment class if possible) | High |
| 15 | overloading_damage_cost | TLD-02, TLD-26 | What is the estimated annual cost of overloading damage? (engine, tire, drivetrain) | High |
| 16 | allotment_forecast_error_cost | TLD-19 | What has been the cost of tire allotment forecast errors? (premium orders, supply gaps) | Medium |
| 17 | total_fleet_maintenance_spend | TLD-46 | What is the total annual fleet maintenance spend? (by equipment class if possible) | High |
| 18 | fleet_age_distribution | TLD-47 | What is the age/hours distribution of the haul truck and shovel fleets? | Medium |
| 19 | annual_rebuild_spend | TLD-47 | What is the annual capital spend on fleet rebuilds? (engines, wheel motors, truck bodies) | Medium |
Maintenance & Operations¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 20 | labor_cost_per_hour (maintenance tech) | TLD-01, TLD-12, TLD-14, TLD-38, TLD-42 | What is the loaded hourly rate for a maintenance technician at Tilden? (wages + benefits + overhead) | Critical — used across 5+ cards |
| 21 | labor_cost_per_hour (supervisor) | TLD-24, TLD-39, TLD-51 | What is the loaded hourly rate for a mine supervisor? | Medium |
| 22 | repairs_per_month (ELLIPS) | TLD-12 | How many maintenance work orders are created per month in ELLIPS? (all types) | High |
| 23 | unplanned_mill_shutdowns_per_year | TLD-03 | How many unplanned concentrator mill shutdowns occurred in the last 12 months? Duration of each? | High |
| 24 | lube_system_failures_per_year | TLD-03 | How many lube system failures per year in the concentrator/pellet plant? | Medium |
| 25 | HPGR_unplanned_downtime_hours | TLD-38 | Total HPGR unplanned downtime hours since installation (April 2023)? | Medium |
| 26 | deferred_maintenance_items | TLD-41 | How many deferred maintenance work orders are in the ELLIPS backlog? What is the estimated cost? | Medium |
Inventory & Procurement¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 27 | total_inventory_value | TLD-35 | What is the total spare parts inventory value at Tilden? (from ELLIPS/Oracle) | High |
| 28 | total_consumable_spend | TLD-13 | What is the annual consumable spend? (tires, liners, reagents, explosives, drill bits combined) | High |
| 29 | boxes_per_day (warehouse receiving) | TLD-35 | How many items/boxes are received per day at the warehouse? | Medium |
| 30 | equipment_purchases_per_year | TLD-48 | How many new pieces of equipment are purchased per year requiring ELLIPS onboarding? | Medium |
Mining Operations & Blast¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 31 | annual_explosive_spend | TLD-05 | What is the annual explosive spend at Tilden? | High |
| 32 | total_tons_hauled_per_year | TLD-04 | What is the total tonnage hauled by the truck fleet per year? | Medium |
| 33 | plan_deviation_events_per_year | TLD-50 | How frequently does mine plan execution deviate from the daily plan? (estimated frequency) | Medium |
Logistics¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 34 | labor_cost_per_hour (logistics/scheduling) | TLD-16 | What is the loaded hourly rate for logistics/scheduling staff? | Medium |
| 35 | wasted_crew_deployments_per_year | TLD-16 | How many train crew deployments per year are wasted due to vessel no-shows or schedule changes? | High |
| 36 | demurrage_cost | TLD-16 | What is the annual demurrage cost? (vessel waiting/delay fees) | Medium |
Safety & HR¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 37 | MSHA_fine_history | TLD-24, TLD-20 | What is Tilden's MSHA fine history? Annual fine spend? | Medium |
| 38 | new_hires_per_year | TLD-14 | How many new hires per year at Tilden? Average onboarding time? | Low |
| 39 | grievance_rate | TLD-52 | How many labor grievances per year? Average cost per grievance? | Low |
Budget & Financial¶
| # | Variable | Needed For | Question to Ask CLF | Priority |
|---|---|---|---|---|
| 40 | total_maintenance_spend (site) | TLD-36, TLD-41 | What is Tilden's total annual maintenance spend? (fleet + fixed plant combined) | High |
| 41 | budget_variance_% | TLD-36 | What is the historical maintenance budget variance? (last 3 years) | Medium |
| 42 | margin_loss_per_off_spec_ton | TLD-06, TLD-09 | What is the margin loss per ton for off-spec pellets? (downgrade pricing vs. spec pricing) | Medium |
Summary: 42 variables needed. 4 Critical (used across many cards), 15 High, 17 Medium, 6 Low priority.