Automotive CAN Signal Mapping Specialist

About OpenTrain

OpenTrain is the #1 platform for starting and growing careers in AI training and data labeling. We connect skilled contributors with projects that teach and refine machine intelligence — and building a profile on OpenTrain is free.

We focus on practical, hands-on tasks that directly shape how AI systems behave. Contributors work remotely, often part-time and flexibly, while gaining experience on real-world datasets and tooling.

About AI training work in automotive

AI training (data collection and annotation) is the human layer behind modern vehicle intelligence — from driver behavior models to diagnostic tools. Projects range from labeling sensor logs to mapping domain knowledge (like which CAN signals matter for a given use case).

This role sits at the intersection of automotive domain expertise and dataset generation: your knowledge of vehicle systems improves data quality and ensures models learn the right signals for real problems.

The role

You will be a contractor who uses our internal proprietary tooling to build a dataset that maps natural-language use case descriptions to the CAN signals to collect for each use case.

Example mapping: "I need to understand how driver uses the sunroof" -> [signal_for_sunroof, signal_for_driving, signal_for_speed, signal_for_ac]. Deliverables are document-format mappings created inside the internal tool.

What you'll do

• Use the provided internal proprietary tooling to create structured mappings from NL use cases to CAN signal lists.
• Translate use-case descriptions into precise signal sets that address the requested analysis or diagnostic goal.
• Apply your automotive/CAN domain knowledge to identify relevant signals, sensors, and contextual signals (e.g., speed, gear, HVAC) per use case.
• Follow project examples, formatting rules, and quality guidelines delivered with the task.

Requirements

Subject-matter expert in automotive CAN signals is mandatory. This project requires real domain knowledge — applicants without automotive/CAN experience should not apply.

• Experience level: Intermediate (this is not an entry-level task).
• Employment type: Contractor (remote; worldwide applicants welcome).
• Compensation: Fixed-price contract — USD 2000 total.
• Data type and label work: DOCUMENT data, label type DATA_COLLECTION, using INTERNAL_PROPRIETARY_TOOLING.
• Preferred (but not mandatory) prior experience: vehicle diagnostics identifying key signals for troubleshooting; vehicle insurance work using driving data to infer premiums; remote vehicle support reading signal data for diagnosis; working with OEM data science teams.

Who should apply

Apply if you have hands-on experience with vehicle CAN architectures, signal naming/interpretation, or working with OEM/telematics data teams. Practical exposure to diagnostics, insurance telematics, or remote support workflows is a strong fit.

This role is ideal for engineers, diagnosticians, or analysts who can reliably translate a business or research use case into the minimal, relevant set of CAN signals to collect.

How the project works

You will be given examples and quality criteria and will produce mappings inside our internal tool. Deliverables are document-format mappings exported or stored in that tooling per project instructions.

As a contractor you will complete the work for the fixed price (USD 2000). Exact timelines, submission format, and acceptance criteria will be shared when the project is assigned.