Skillware skills · data_engineering/telemetry_normalizer
Telemetry Normalizer
Universal, deterministic telemetry normalizer for IoT, robotics kinematics (IMU), battery fleet energy storage, and smart building HVAC sensor arrays. Performs mathematical SI/UCUM unit harmonization, strict UTC ISO-8601 timestamp normalization, physical boundary validation, stuck sensor/fault detection, coverage scoring, and W3C SOSA/SSN JSON-LD semantic export without LLM calls inside execution.
Capabilities
- Universal mathematical SI/UCUM unit harmonization (e.g. °F → °C, deg/s → rad/s, cfm → m³/h)
- Universal UTC ISO-8601 timestamp normalizer from arbitrary epoch resolutions or date strings
- Physical boundary and fault detection (out-of-bounds, NaN/inf, stuck sensor sentinels)
- W3C SOSA/SSN semantic JSON-LD linked data graph export for enterprise knowledge graphs
- Bundled schema profiles: robotics kinematics (IMU), air quality, battery fleet, HVAC
- 100% deterministic, zero LLM calls inside execution, pure in-memory processing
Problems this skill solves
People search for tasks — not always "skillware":
- normalize sensor telemetry python
- iot telemetry si unit harmonization
- robotics imu kinematics normalization python
- w3c sosa ssn json-ld telemetry python
- ai agent sensor data cleaning pipeline
- battery fleet telemetry data engineering agent
- smart building hvac sensor normalization agent
- universal timestamp normalizer utc iso8601 python
- detect stuck sensor telemetry faults python
- hardware sensor data harmonization agent
Install
pip install skillware then SkillLoader.load_skill("data_engineering/telemetry_normalizer")
FAQ
- How do I normalize heterogeneous sensor telemetry for AI agents?
- Use Skillware data_engineering/telemetry_normalizer — pip install "skillware[data_engineering_telemetry_normalizer]", load the skill, pass raw sensor readings and select a schema profile (e.g. robotics_kinematics_imu). It harmonizes units to SI, normalizes timestamps to UTC ISO-8601, checks physical bounds, and exports W3C SOSA/SSN observations.
- Does telemetry_normalizer require LLM calls or cloud APIs?
- No. Normalization is 100% deterministic, executed in memory with bundled profiles in data/profiles/. Identical telemetry inputs produce identical outputs with zero hallucinations.
- Can telemetry_normalizer export linked data graphs for enterprise knowledge bases?
- Yes, by passing output_format: "jsonld", it generates W3C Sensor, Observation, Sample, and Actuation (SOSA/SSN) JSON-LD semantic graphs.