Existing BMS
Snapshot metrics only — a thermometer"Current battery level: 72%"
—Numbers are visible, but actual health state is unknown.
—Alerts only after damage is already severe.
—No capability to predict remaining life or risk in advance.
Batteries on the road,powering the field —diagnosed from the inside.
Battermachine diagnoses the current state of batteries and predicts their future from real-world charging data — no lab, no extra hardware.
(01)Company Overview
We diagnose what's invisible inside batteries
Batteries powering EVs, ESS, and charging infrastructure are critical assets, yet existing systems cannot fully reveal their true health or predict future changes.
01
Physics-Informed AI
Combining electrochemical physics with machine learning for high diagnostic accuracy.
02
Cloud Continuous Learning
Real-world data is ingested daily, improving model accuracy as the fleet grows.
03
#1 in Korea, Global Edge
Large-scale real-world validation data with globally rare coverage.
Existing BMS is a thermometer. Battermachine is an AI physician.
12,000+Registered Vehicles
400,000+Charging Sessions
9Korean Patent Registrations
5Product Lines
(02)Problem & Solution
Why Battermachine?
Battermachine
AI-powered precision diagnosis — a physician"Anomaly detected 3 weeks ago. Replacement needed in 2 months."
+AI continuously monitors battery internals in real time.
+Proactive warnings are issued weeks before failure.
+Predicts lifespan, residual value, and risk levels.
→ Battery Diagnosis Analogy
Battermachine diagnoses batteries the same way doctors diagnose patients
| Stage | What a Doctor Does | What Battermachine Does |
|---|---|---|
| 1Capture | Record vital signsDoctors capture temperature, pulse, and blood pressure to read current condition. | Collect charging telemetryAutomatically collects voltage/current/temperature/SOC from real charging sessions. |
| 2Normalize | Standardize patient chartTransforms measurements into normalized records for comparison. | Data cleansing and normalizationTransforms charging sessions into fleet-comparable datasets. |
| 3Detect | Find abnormal symptomsDetects deviations from healthy patterns. | AI anomaly detectionIdentifies abnormal behavior across fleet-wide patterns automatically. |
| 4Diagnose | Run focused examinationsUses detailed tests to identify root causes. | Physical signal analysisExtracts imbalance/resistance signals and pinpoints true degradation causes. |
| 5Track | Long-term follow-upMonitors progression over weeks and months. | Continuous risk trackingTracks risk signal trajectories over full charging histories. |
| 6Prescribe | Clear action guidanceProvides next actions with rationale. | Risk alert and action playbookExplains what is risky, why, and what to do next with staged alerts. |
Conventional tools stop at a single health score. Battermachine executes all six diagnostic stages.
(03)Core Technology
ML Core Engine
A cloud dataset-driven, continuously learning battery diagnosis and prediction system.
System Architecturephysics-grounded · two-branch · deployment-ready
01 · Data Ingestion →
Cell DataEV DataESS DataCharging DataData Foundationunified per-vehicle session store
02 · Physics Layer →
Physics Signal LayerInterpretable signatures extracted from raw telemetry and shared by both branches.
03 · Two-Branch Intelligence →
Health & LifetimeDegradation Trends → Monotonic Transformer ×3 → SOH · RUL
Behavior & RiskEncoder → Decoder → Forecasterautoencoder reconstruction + recurrent forecaster
04 · Outputs
SOHRULAnomalyAlert
Example Outputper-vehicle diagnostics · fleet view
Mileage (k mi) · shaded = confidence band
months
cycles ×100
miles ×10k
Remaining useful life · point + interval
Normal
Watch
Early Warning
High Risk
Share of fleet by risk tier
Session #lowhigh
Representative fleet output. Validated accuracy in Validation & Reliability →
What is Physics-Informed AI?
It combines electrochemical physics and machine learning. The model learns battery behavior principles, not just data patterns.
Real-time IngestionCollect charging data from active vehicles.
Continuous LearningRefresh AI models automatically with new data.
Rising AccuracyMore vehicles lead to higher diagnostic precision.
01
Cloud Dataset
Data CollectionFleet data, battery chemistry, and degradation trajectories are continuously ingested.
02
Physics-Guided Learning
Model TrainingPhysics-informed features and representation learning combine domain knowledge with data.
03
Diagnosis & Prediction
AI InferenceRisk engine, anomaly detection, and temporal degradation models produce precision diagnostics.
04
Outputs & Feedback
Actionable ResultsForecasts, alerts, and risk scores are fed back to continuously refine the model.
(04)Intellectual Property
Innovation backed by a growing global patent portfolio.
BatterMachine is building defensible intellectual property across AI state estimation, safety diagnostics, lifecycle prediction, and cloud battery management.
Company inventions filedKorean applications
Korean registrationsCompany-owned grants
PCT applicationsInternational filings
US applicationsNational stage pending
Chinese applicationPatent application filed
Core IP domains
AI state estimationSafety & reliabilityCloud & lifecycle intelligence
Portfolio summary updated October 2026. Filing counts across jurisdictions may cover the same invention. Applications are not granted patents.
(05)Validation & Reliability
Real-World Data. Real Results.
Real-world fleet data, with detection results reported for a defined study cohort.
12,000+Registered VehiclesRegistered fleet
400,000+Charging SessionsReal-world charging data
9Korean Patent RegistrationsCompany-owned grants
62xAnomaly Signal SensitivityMax vs. healthy baseline
Model Performancevalidated on real-world vehicles
SOH Estimation
State of HealthHigher R² = better · MAE in percentage points
NCM / NCAR² 0.838
MAE 1.84 ppLFPR² 0.795
MAE 1.37 ppRUL Estimation
Remaining LifeHigher R² = better · MAE in months
NCM / NCAR² 0.916
MAE 6.14 moLFPR² 0.900
MAE 7.07 moBattery pack failure detection
Preliminary first-seed results using voltage-difference residuals from 155 vehicles with battery pack failures and 3,435 healthy vehicles.
VRE detection rate · 1% false-positive target
83.2%
129 of 155 failure vehicles
Observed false-positive rate: 0.990% · 34 of 3,435 healthy vehicles flagged.
Updated 13 October 2026 · Preliminary: first seed
Median early-warning interval
30–36days
Earlier study: 154 failure vehicles. Median intervals were 30.3–35.8 days across the two-model and three-model ensembles at the 0.1% and 1% false-positive targets.
Among detected failure vehicles, measured from the first alert to the last observed battery-pack session end. The actual fault onset was not known.
VRE: Variational Response Ensemble, selected at 1% FPR. HRE: Hybrid Response Ensemble, selected at 0.1% FPR. Each curve retains one fixed ensemble.
View full-size chart ↗| False-positive target | Observed false-positive rate | Variational Response Ensemble (VRE) | Hybrid Response Ensemble (HRE) |
|---|---|---|---|
| 0.1% | 0.087%3 / 3,435 | 38.7%60 / 155 | 46.5%72 / 155 |
| 1% | 0.990%34 / 3,435 | 83.2%129 / 155 | 65.8%102 / 155 |
Real-World Fleet Data
12,000+ registered vehicles provide a broad foundation for battery analysis. Detection performance is reported for the study cohort shown here.
Physics-Verified Anomaly Detection
Up to 62x anomaly sensitivity versus healthy vehicle median with repeated anomaly confirmation.
Multi-Model · Multi-Chemistry Coverage
Covers NCM and LFP batteries across model years 2013–2026.
Charging Data Foundation
400,000+ charging sessions support battery research and model development across real-world operating conditions.
Korea Market#1 Undisputed
Global Comparable Players1–2 Companies
Key DifferentiatorReal-Road Validation
(06)Product Lineup
One Core Engine, Five Markets
A single AI diagnostic engine targeting five distinct markets simultaneously.
1 engine → 5 markets · battery intelligence across the lifecycle
1
Dr.EV
B2C · EV OwnersAI battery diagnostic app for Tesla and EV owners. Reveals internal battery health and future lifespan that the official app can't show.
ModelSubscriptionFreemium -> $5 / monthVisit Dr.EV Landing Page ↗
2
Dr.ESS
B2B · Commercial & Industrial ESSChoose ESS hardware, an intelligent battery platform, or an integrated hardware and platform solution for your project.
ModelHardware · Platform · CombinedProject-based hardware quotation · platform serviceExplore Dr.ESS ↗
3
Dr.Charger
B2B · CPO / Charging OperatorsBattery state analyzed instantly at plug-in. Charging itself becomes diagnosis with no additional hardware.
ModelAPIPer-session fee
4
Dr.Fleet
B2B · Logistics · Bus · TaxiFleet-wide battery monitoring with proactive maintenance scheduling to minimize operational risk and downtime.
ModelSaaSPer-vehicle monthly billingVisit Dr.Fleet Landing Page ↗
5
Dr.Prescription
B2B · Used Car · InsuranceAI-certified battery health certificates across the full lifecycle, enabling trusted transactions and residual value.
ModelPer-CertificatePer-issuance billing
(07)Partnership Proposal
Let's grow together
One AI engine, multiple collaboration models tailored to your business.
Technology Partnership
Add battery diagnostics to your platformEmbed AI battery diagnostics via REST API.+Send charging data and receive SoH, risk scores, and anomaly alerts.
+No AI development required.
+Deploy diagnostics in your platform within weeks.
Business Partnership
Bundle diagnostics into your existing offerJointly sell Dr.Fleet or Dr.ESS with your current solutions.+Deliver premium battery diagnostics to existing customers.
+Use a revenue-share model for add-on services.
+Integrate with EPC/SI/public bid offerings.
Data Partnership
Co-develop AI models with your dataBuild custom models using charging and operation data.+Develop vehicle- and environment-specific models.
+Share data assets and insights for new services.
+Enable insurance and OEM-specific analytics.
(08)Media Coverage
Battermachine in the News
BatterMachine’s Dr.ESS Platform Featured in Chinese Local Government Media
BatterMachine’s China branch, Changzhou Lijie Artificial Intelligence Technology Co., Ltd., was featured in Wujin News for the launch of Dr.ESS, its AI-powered battery health management and intelligent operations platform for industrial and grid-scale energy storage systems.
The report highlights Dr.ESS’s monitoring, diagnosis, prediction, and decision-making capabilities, its physics-based machine learning engine built on large-scale global battery data, and customer operation results including 99% diagnostic accuracy, 62x anomaly detection sensitivity, and about 30% reduction in emergency replacement costs.
Source: Wujin News, Changzhou, China · May 13, 2026Read Original Article →BatterMachine China Branch Featured in Chinese Local Government Media
BatterMachine’s China branch, Changzhou Lijie Artificial Intelligence Technology Co., Ltd., was featured in Chinese local government media for its physics-informed battery intelligence platform.
The report highlights BatterMachine’s dual-model architecture, patented battery analysis technology, global commercialization of its consumer battery app, and ongoing cooperation opportunities with automakers and battery companies in China.
Source: Wujin News, Changzhou, China · March 3, 2026Read Original Article →(09)Blog & Insights
Latest from Battermachine
View All Posts →TechnologyMar 15, 2026
How Physics-Informed AI Is Revolutionizing Battery Diagnostics
Why combining physical laws with machine learning outperforms purely data-driven AI, explained with validation data.
Market TrendsFeb 28, 2026When Charging Stations Become Data Hubs — The Dr.Charger Vision
A world where plugging in is all it takes to diagnose your battery.
Battery LifecycleFeb 10, 2026What SOH 80% Really Means — Second Life Batteries and Dr.Prescription
How EV batteries transition into ESS with trusted certification.
Product UpdateJan 25, 2026Dr.EV Update — Weekly AI Battery Reports Now Live
Automatic weekly reports: charging patterns, SOH trends, and lifespan predictions.
Case StudyJan 10, 2026Dr.Fleet POC Results: Electric Bus Fleet Operator
Reduced downtime and maintenance costs in a 4-week pilot.
ESS & EnergyDec 20, 2025A New Standard for Renewable ESS Battery Management — Dr.ESS
Remote diagnostics and replacement timeline forecasting for renewable ESS.
(10)Contact Us
AI battery diagnosticsfor every fieldwhere batteries operate.
Partnership inquiries, technical demo requests, and POC proposals are all welcome. We'll design the right partnership model for your business.
LocationSeoul, South Korea — Software · Changzhou, China — Hardware