Review Methodology

📅 Last reviewed: July 3, 2026

Our app and hardware reviews are the result of systematic, real-world testing rather than spec-sheet summarisation. This document explains exactly how we evaluate BMS monitoring applications and the hardware they work with.

Testing Hardware Setup

All app reviews are conducted against physical BMS hardware running actual battery packs. Our standard test configuration uses:

Testing Protocol

Phase 1: Connection Testing

We evaluate Bluetooth discovery time, pairing reliability across multiple phone models, range in open air and within metal enclosures, and reconnection behaviour after interruptions.

Phase 2: Data Accuracy

We compare the app's displayed values against our calibrated reference instruments. Acceptable accuracy thresholds: voltage ±0.05V per cell, current ±2% of full scale, temperature ±1°C, SOC ±5% (after full calibration cycle).

Phase 3: Alert System Testing

We deliberately induce each alert condition in a controlled, safe environment and verify the app responds correctly. For over-temperature alerts, we use a heating element near (not on) a thermistor. For over-voltage, we use a bench power supply on an isolated cell.

Phase 4: Data Logging

We run a complete 8-hour logging session during a discharge-recharge cycle and verify log completeness, timestamp accuracy, and export integrity.

Phase 5: User Experience

We assess UI intuitiveness, navigation depth, settings accessibility, and language clarity. We recruit 2–3 non-technical users (representative of the target user base) to complete common tasks without guidance and document where confusion arises.

Scoring System

Our verdict score (0–100) is calculated from five weighted categories:

CategoryWeightWhat We Evaluate
Data Accuracy30%Voltage, current, temp, SOC vs. reference instruments
Feature Completeness25%Cell monitoring, alerts, logging, export, config
Reliability20%Connection stability, crash rate, background operation
Security15%Authentication, BLE security, permission use
User Experience10%Interface clarity, navigation, language accessibility

Scores are assigned on each criterion by two independent revieweDK and averaged. A single category falling below 40% can cause the overall score to be capped, regardless of performance in other areas.

Independence & Conflicts of Interest

We purchase or independently source all test hardware. We do not accept free hardware, software licences, or any other consideration from manufacturers whose products we review. RevieweDK are required to disclose any prior business relationship with the company being reviewed before being assigned that review.