Open incident dataset · 341 reports

Tracking reported ICCU incidents across E-GMP electric vehicles

This site aggregates publicly submitted reports about suspected ICCU- related incidents in E-GMP platform vehicles into a single anonymized dataset. The goal is to make reported model, mileage, build-date, and recall patterns easier to review without relying on scattered forum posts or screenshots.

Independent community project. Not affiliated with Hyundai Motor Group, Kia, or Genesis.

Total reports These are all published rows from the moderated export, including both form submissions and manually added records. 341
Direct submissions Published reports submitted through the Google Form. 10
Avg km at reported incident39,146
Last updated The site refreshes on new pushes and on the daily GitHub Pages refresh job. 2026-03-28

What is the ICCU issue?

The Integrated Charging Control Unit is an important power-electronics component in these vehicles. The summary below reflects general system function and community hypotheses, not a manufacturer-confirmed root cause analysis.

What the ICCU does

The ICCU converts high-voltage DC from the main battery to 12 V to keep the auxiliary battery charged. It also enables Vehicle-to-Load (V2L) output. Without it, every low-voltage system in the car depends on that power path, including lights, displays, door locks, and some support systems.

What owners report

A commonly discussed theory is that some reported incidents may involve electrical stress, heat, charging-related conditions, or other internal failure modes. In owner reports, these incidents are often associated with dashboard warnings, reduced power, "turtle mode", charging problems, or in some cases an undriveable vehicle. Those reports do not by themselves establish a single cause.

🚗

Which vehicles appear in reports

Several E-GMP models appear in public discussions about suspected ICCU-related incidents, including Ioniq 5, Ioniq 6, EV6, GV60, GV70 Electrified, and G80 Electrified. Public recall actions and service campaigns have also been discussed, but this site does not claim to prove which vehicles share the same underlying issue or whether any specific update resolves it completely.

Why does this tracker exist?

Reports of possible ICCU-related incidents are scattered across forum threads, Reddit posts, and social media, which makes broader pattern review difficult. This project centralizes anonymized reports into a single, open dataset so people can explore possible correlations, such as build months that appear more often, mileage ranges where reports seem to cluster, or whether incidents are still being reported after specific recall actions. The dataset is intended as a public research aid for owners, journalists, consumer advocates, and safety investigators.

Live analytics

Filter and explore all published incident reports in real time. The statistics below summarize submitted reports and should be read as descriptive observations, not proof of causation.

⚙ Filters
Incidents in view341
Avg km at reported incident39,146273 with km data
Possible repeat incidents21% of selection
After recall 41D0334714% of selection

Incident table

All published reports, filtered by your current selection. Click column headers to sort, use column menus to filter. Only anonymized, normalized fields are included.

How to contribute

Start with the report form. Submit the incident details once, and the moderation workflow handles the rest before publication.

Submission flow

Send the incident details, then we clean them up

The form stays short. Submit the vehicle and incident details once, and the moderation process takes care of the rest.

1

Fill out the short form

Click "Report a suspected ICCU incident" and fill in the Google Form. You'll be asked for the vehicle model, region, approximate build date, odometer reading at the time of the reported incident, and whether the 41D033 recall was done beforehand.

2

We review and normalize

Submissions land in a private moderation sheet. We verify plausibility, normalize model names and dates, and flag duplicates. Approval means the report is suitable for the public dataset, not that the underlying theory has been technically proven.

3

It appears on this page

When the site rebuilds, the public CSV export is converted to a JSON snapshot and the analytics update automatically. Your report becomes part of the live charts and downloadable dataset.

Contribute on GitHub

Help improve the tracker, data pipeline, and documentation

This project is open source. Use the repository links below to report issues, propose documentation improvements, or send patches.

Privacy and data handling

The tracker is designed to publish useful incident patterns without publishing personally identifying information.

What is stored privately

Raw submissions enter a moderation sheet first. That working copy may include enough detail to validate the report, detect duplicates, and standardize dates, regions, and model names before publication.

What is published publicly

The public dataset only exposes normalized, anonymized fields used for analysis. Names, email addresses, screenshots, raw submission IDs, and other direct identifiers are not published. Moderated public comments may be published when maintainers decide they are safe to expose. The public table and exports are limited to non-identifying fields needed for analysis.

Corrections and removal

If a published row needs a correction or should be removed, open an issue in the project repository or contact the maintainers through the submission workflow so the moderation sheet can be updated before the next snapshot.

Methodology and dataset notes

This page should be treated as a descriptive incident tracker, not a failure-rate study or proof of causation.

Publication criteria

Rows are published only after moderation. The review checks whether the report is coherent enough to normalize and whether it belongs in the public ICCU incident dataset.

Normalization and deduping

Model labels, recall status, build dates, and reported failure dates are normalized into a common format. Repeat incidents are flagged so the same vehicle can be studied without counting every report as an entirely unrelated case.

Field definitions

The published view focuses on model, region, build date, odometer at failure, failure date precision, recall status, repeat-incident flagging, moderated public comments, and coarse VIN sequence buckets. Missing or uncertain values are shown as unknown rather than guessed.

Open data

Download the current moderated snapshot

Use the full snapshot if you want to audit the current public data without applying page filters first.