JD Plan
Battery diagnostics · Deployment and end-of-life decisions
Battery-by-battery decisions based on predicted remaining life.At deployment, it scores ESS fitness and economics. At end of life, it routes each battery to reuse, recycling, or disposal.
01Problem
Two decision points, one underlying question
Both come down to one question: how much longer will this battery last?Today, most answers rest on a single test.
BOL
No standard for deploying reused batteries
Makers, specs, and wear vary widely.There's little evidence to say whether a given battery belongs in an ESS.
EOL
Uncertain end-of-life routing
Whether a retired battery is reused, recycled, or scrapped hinges on a single retest.
Impact
The cost of playing it safe
Without evidence, reusable batteries get recycled.That means lost reuse volume and lost revenue.
02Process
Four-step diagnosis
From data upload to dispatch decision.Results for every battery.
- 01
Upload BMS data and history
Cell-level time series and operating records
- 02
Predict SOH and RUL
Physics × AI ensemble inference
- 03
Decide per battery
Fitness, routing, and a report
- 04
Deploy to ESS or ship to recycling
Sorted into reuse, recycle, or dispose
03Features
Key capabilities
- 01
Go / no-go decisions
Determines whether each battery is fit for ESS based on predicted remaining life.Uses cell-level BMS time series and operating history as inputs.
- Battery-level decisions
- Thresholds configured to your criteria
- Same model for new and reused batteries
- 02
Economics forecast
Estimates service life and returns for batteries cleared for ESS.Based on when the predicted degradation curve hits the operating floor.
- Expected service life per battery
- Return metrics for investment decisions
- 03
End-of-life routing
Routes batteries based on remaining capacity and operating history.With history from JDMS or JD EMS, no retesting is needed.
- Reuse — redeploy at an ESS site (BOL)
- Recycle — black mass and material recovery
- Dispose
- 04
Decision reports
Issues a report for every battery with the decision and the evidence behind it.Includes predictions, input data coverage, and rationale.
- Results by battery ID
- SOH and RUL predictions with supporting evidence
04Status
Built and in field validation
- 1/5
- SOH prediction error vs. conventional methods
- PoC
- Joint validation of the AI life model with LS ELECTRIC
- MoU
- Co-developing reuse ESS with LG Energy Solution
MAE 0.62pp vs 3.24pp · n=1,428
Model validation
BOL · ESS
05Lifecycle
How data moves between stages
BOL · Deployment
JD Plan
Deployment decision · Economics
MOL · Operation
JDMS · JD EMS
Builds operating history
EOL · End of life
JD Plan
End-of-life routing
- Input
- Operating history from JDMS and JD EMS, plus cell-level BMS time series
- Output
- Reuse → into JD EMS / Recycle → into JD Works