Value & Performance Ledger
Fleet Impact Since Deployment
Net Value Delivered · This Month
€181,400 vs Unoptimised Baseline iProduction protected plus energy spend avoided, benchmarked against the same fleet running without the optimiser. Battery and asset-life gains accrue on top over the replacement horizon.
Fleet Availability
84%
iPercentage of the period the active fleet was held above the production line's minimum AGV requirement.
Availability
47
Depletion Events Prevented
iCharge-driven shortfalls that would have pulled the active fleet below the production line's minimum AGV requirement.
Reliability
11
Unplanned Breakdowns Avoided
iDeveloping faults detected by condition monitoring and resolved before causing a mid-shift failure.
Energy
€7,400
Energy Cost Avoided
iSavings from tariff-aware load shifting: charging moved into off-peak windows reduces per-kWh cost and lowers site demand peaks.
Availability
Sustaining Fleet Availability
iThe optimiser continuously forecasts fleet state of charge and pre-empts availability shortfalls – scheduling charge top-ups before too many AGVs deplete simultaneously. Vehicle supply to the line stays uninterrupted, without the manual firefighting it would otherwise demand.
Depletion Events Prevented
iAGVs that would have hit charge exhaustion mid-task without the optimiser intervening.
47
Estimated value protected
iProduction downtime avoided. Based on €750/hr agreed line-downtime cost × 1 hr per depletion event.
€174,000
Fleet Performance: Before vs. After
+6pp
Availability
78% → 84%
+1pp
Charging
14% → 15%
−7pp
Out of action
8% → 1%
Fleet Availability: Optimised vs. Baseline
iThe shaded band shows availability headroom added by the optimiser. Bars show depletion events that would have occurred in the unoptimised baseline.
Optimised
Baseline
Depletion Events Prevented
Reliability
Condition Monitoring and Breakdown Prevention
iEach AGV's power draw, charge behaviour, and thermals are continuously profiled. Developing anomalies are surfaced for maintenance while the vehicle is still in service, converting unplanned mid-shift failures into scheduled interventions.
Condition monitoring · this period
iAGVs flagged for abnormal charge behaviour before causing a mid-shift failure. Not every detection results in a breakdown: the conversion rate reflects the estimated share that would have failed without intervention, configurable in dashboard settings.
31
Anomalies flagged
11
Breakdowns avoided
Estimated value · this period
€665
11 breakdowns avoided
× €300 per breakdown (agreed rate)
Each breakdown avoided is estimated to save one technician call-out. Time saved is based on the average response-plus-repair duration observed for this vehicle type. The labour rate is the fully-loaded cost including overhead, configurable in dashboard settings. Battery and asset-life benefits from reduced charge stress are captured separately in Battery life.
Energy
Tariff-Aware Charging and Load Shifting
iCharge sessions are scheduled against live tariff and grid signals, shifting load into off-peak windows to reduce energy cost and site demand peaks – without affecting fleet availability.
Tariff Saving
ikWh shifted from peak hours (06:00–22:00 at €0.22/kWh) to off-peak hours (22:00–06:00 at €0.14/kWh). Total energy consumed is unchanged – only the timing of charging changes. Rates are illustrative; update with site-specific tariffs in dashboard settings.
€4,400
Load shifted to off-peak tariff windows
Avg. Rate Before
€0.22/kWh
Avg. Rate After
€0.14/kWh
Demand Charge Saving
iIndustrial tariffs include a demand charge based on peak kW drawn in any 15-minute billing window. Scheduling AGV charging away from site demand peaks reduces this figure directly.
€1,600
Fixed savings per month
Peak kW Avoided
32 kW
Demand charge rate
€50/kW
Charging Load Profile
iEach row = one hour of the day (00h–23h). Each column = a time period (day view: hours, week view: days, month view: days of the month). Colour intensity shows how many AGVs were charging. Amber markers on the Y-axis indicate high-tariff hours; red markers indicate site demand-peak hours.
Few chargers
Many chargers
High tariff hour
Demand peak hour
Before
After
Battery life
Battery Health and Asset Life Extension
iCharge optimisation reduces battery stress by limiting deep discharges. Fewer low-SOC events extend battery service life and defer fleet replacement CapEx.
Average minimum SOC per shift
iAverage lowest state of charge reached before the AGV begins charging, per vehicle per shift. Higher is healthier – deep discharges accelerate cell degradation.
38%
Before
18%
After
38%
Deep discharge events
iAGVs reaching below 15% SOC this period. Below 15% is the zone where lithium cell stress accelerates significantly, shortening overall battery lifespan.
2
Before
14
After
2
Projected life extension
iEstimated additional service life per battery, derived from the reduction in deep discharge frequency. Batteries are retired at 80% state of health – fewer stress events push that threshold later.
+5.5 mo
Batteries in fleet
55
CapEx deferred
~€150k
Minimum SOC Distribution: Before vs. After
iHow deeply batteries discharged before the next charge cycle. A distribution shifted right (higher minimum SOC) indicates reduced cell stress and longer battery life. The shaded zone marks the high-stress region below 15% SOC.
Before
After
High stress zone (<15%)
Confidence
Model Accuracy and Prediction Confidence
iValue estimates across this report depend on accurate predictions of when each AGV will reach its critical SOC threshold. This section shows how closely those predictions matched observed outcomes.
Mean Prediction Error
iAverage absolute difference between predicted and actual time-to-critical SOC, across all AGVs this period. Lower is better; under 5 minutes is the operational threshold for reliable charge scheduling.
4.2 min
Prediction bias
iMean signed error (actual minus predicted). A value near zero indicates the model is neither consistently early nor late. A small positive bias means the model predicts critical SOC slightly sooner than observed, providing a conservative safety buffer.
+0.3 min
Prediction Error Distribution
iEach bar shows the number of predictions in that error band (minutes early or late). A tight distribution centred near zero indicates the model is both accurate and unbiased: errors are small and do not skew consistently in one direction.