Upload sample dataset files

QRaco Fusion Sample is a public sample dataset for motion, gesture, and sensor-fusion research.

It contains mobile sensor data captured from an Android device and is meant to demonstrate the structure, format, and quality of the full QRaco Fusion dataset.

Included files:
- records_part_001.jsonl.gz
- metadata.json
- schema.json
- README.txt
- integrity.json

This sample is provided for research, evaluation, and preview purposes only. The full commercial package is available separately.

Sensors represented in the sample may include accelerometer, gyroscope, magnetometer, light, proximity, and battery temperature, depending on the recording session.

README.txt

@@ -0,0 +1,18 @@
+QRaco Physical AI Recorder dataset
+
+Primary data: records_part_*.jsonl.gz
+Storage profile: compact primary sensors, source Float32 precision preserved.
+Derived Android streams are omitted because gravity and linear acceleration can be reconstructed offline.
+Protocol and device metadata: metadata.json
+Field contract: schema.json
+Compact integrity chain: blocks.jsonl
+File checksums: integrity.json (created when ZIP is exported)
+Auto-detected rare events are heuristic candidates, not ground truth.
+Confirm important events with manual markers and document the collection protocol.
+
+Parquet conversion on a PC:
+  1. Install Python and run: pip install pyarrow
+  2. Put convert_to_parquet.py beside the JSONL.GZ parts.
+  3. Run: python convert_to_parquet.py
+
+The converter is streaming and writes records.parquet with ZSTD compression.

integrity.json

@@ -0,0 +1,36 @@
+{
+  "algorithm": "SHA-256",
+  "generated_at_epoch_ms": 1783733275208,
+  "files": [
+    {
+      "name": "README.txt",
+      "size_bytes": 835,
+      "sha256": "13fe353991ed465dde47bda855fd7b86ecfe26b1238644cb3f36bc0326b1f2d1"
+    },
+    {
+      "name": "blocks.jsonl",
+      "size_bytes": 5853,
+      "sha256": "c2ddbc724afbad2705e5777ed21f6ba04131bc648d8e588abed4cd861f9a6efe"
+    },
+    {
+      "name": "convert_to_parquet.py",
+      "size_bytes": 1043,
+      "sha256": "747795d9ab7ffa5b263c2eb4ffc6a8cc52fe21382a9a1ab0ac5d3139b80f5eb8"
+    },
+    {
+      "name": "metadata.json",
+      "size_bytes": 7509,
+      "sha256": "7e3bbb4e50aac255a839b8afdd860e097c187317b07f382ab729705c32b86113"
+    },
+    {
+      "name": "records_part_001.jsonl.gz",
+      "size_bytes": 2215149,
+      "sha256": "4c36b07deada6a4503b3d862b901f0322e172fc00771c7dd9247bd83fb7f8b16"
+    },
+    {
+      "name": "schema.json",
+      "size_bytes": 1443,
+      "sha256": "3e0677e9d65de6097b94da2c63a5fcec9237b47d8fc4c74c8d3e0f967d9f9b8f"
+    }
+  ]
+}

metadata.json

@@ -0,0 +1,253 @@
+{
+  "schema_version": "1.0.0",
+  "session_id": "20260711_005336_1817813b",
+  "dataset_name": "Physical AI field capture",
+  "campaign_id": "QRPAI-001",
+  "protocol_id": "PROTO-001",
+  "contributor_code": "P-001",
+  "known_activity": "pocket_movement",
+  "custom_activity": "",
+  "task_outcome": "unknown",
+  "environment": "interior",
+  "surface": "necunoscuta",
+  "equipment": "telefon purtat",
+  "conditions": "",
+  "notes": "",
+  "started_at_epoch_ms": 1783731216220,
+  "ended_at_epoch_ms": 1783733273408,
+  "duration_ms": 2057188,
+  "timezone_id": "Europe\/London",
+  "timezone_offset_minutes": 60,
+  "locale": "en-GB",
+  "record_count": 65640,
+  "data_part_count": 1,
+  "integrity_block_count": 17,
+  "integrity_chain_head_sha256": "c53820d3d1fe0a8f20dfd27ed6122c6164330936a353600b188e4f299e99bddf",
+  "integrity_block_size_records": 4096,
+  "dropped_event_count": 0,
+  "requested_sampling_mode": "SENSOR_DELAY_FASTEST_no_batching",
+  "storage_profile": "compact_primary_sensors_float32_v1",
+  "numeric_encoding": "Shortest decimal representation that round-trips to source Float32; location remains Float64",
+  "calibration_window_ms": 5000,
+  "video_enabled": false,
+  "video_file": null,
+  "location_enabled": false,
+  "privacy_note": "Location is opt-in. No name, email, advertising ID, serial number, microphone audio or account data is collected.",
+  "consent_granted": true,
+  "consent_text_version": "QRPAI-CONSENT-1.0",
+  "consent_recorded_at_epoch_ms": 1783731216220,
+  "device_manufacturer": "HONOR",
+  "device_model": "MBH-N49",
+  "device_product": "MBH-N49EEA",
+  "device_hardware": "qcom",
+  "android_release": "16",
+  "android_sdk": 36,
+  "app_version": "1.3.0",
+  "sensor_descriptors": [
+    {
+      "type": 1,
+      "key": "accelerometer",
+      "name": "lsm6dso Accelerometer Non-wakeup",
+      "vendor": "STMicro",
+      "version": 143119,
+      "resolution": 0.0047856453,
+      "maximum_range": 156.9064,
+      "minimum_delay_us": 2404,
+      "maximum_delay_us": 1000000,
+      "power_ma": 0.17,
+      "reporting_mode": 0,
+      "wake_up": false
+    },
+    {
+      "type": 16,
+      "key": "gyroscope_uncalibrated",
+      "name": "lsm6dso Gyroscope-Uncalibrated Non-wakeup",
+      "vendor": "STMicro",
+      "version": 143119,
+      "resolution": 0.0012217296,
+      "maximum_range": 34.906036,
+      "minimum_delay_us": 2404,
+      "maximum_delay_us": 1000000,
+      "power_ma": 0.55,
+      "reporting_mode": 0,
+      "wake_up": false
+    },
+    {
+      "type": 14,
+      "key": "magnetometer_uncalibrated",
+      "name": "mmc56x3x Magnetometer_Uncalibrated Non-wakeup",
+      "vendor": "memsic",
+      "version": 10000,
+      "resolution": 0.0976,
+      "maximum_range": 3000.0288,
+      "minimum_delay_us": 10000,
+      "maximum_delay_us": 1000000,
+      "power_ma": 1,
+      "reporting_mode": 0,
+      "wake_up": false
+    },
+    {
+      "type": 11,
+      "key": "rotation_vector",
+      "name": "FUSION Rotation Vector  Non-wakeup",
+      "vendor": "akm",
+      "version": 1289,
+      "resolution": 0.01,
+      "maximum_range": 1,
+      "minimum_delay_us": 10000,
+      "maximum_delay_us": 10000,
+      "power_ma": 0.001,
+      "reporting_mode": 0,
+      "wake_up": false
+    },
+    {
+      "type": 5,
+      "key": "light",
+      "name": "TCS3720_1ALSPRX Ambient Light Sensor Non-wakeup",
+      "vendor": "ams AG",
+      "version": 66048,
+      "resolution": 1,
+      "maximum_range": 20000,
+      "minimum_delay_us": 0,
+      "maximum_delay_us": 0,
+      "power_ma": 0.38,
+      "reporting_mode": 1,
+      "wake_up": false
+    },
+    {
+      "type": 8,
+      "key": "proximity",
+      "name": "TCS3720_1ALSPRX Proximity Sensor Wakeup",
+      "vendor": "ams AG",
+      "version": 66048,
+      "resolution": 1,
+      "maximum_range": 5,
+      "minimum_delay_us": 0,
+      "maximum_delay_us": 0,
+      "power_ma": 0.625,
+      "reporting_mode": 1,
+      "wake_up": true
+    },
+    {
+      "type": 6,
+      "key": "pressure",
+      "name": "goertek_spl07 Pressure Sensor Non-wakeup",
+      "vendor": "Goertek",
+      "version": 4864,
+      "resolution": 0.01,
+      "maximum_range": 1100,
+      "minimum_delay_us": 31250,
+      "maximum_delay_us": 1000000,
+      "power_ma": 0.006,
+      "reporting_mode": 0,
+      "wake_up": false
+    }
+  ],
+  "measured_sensor_statistics": {
+    "accelerometer": {
+      "count": 35836,
+      "measured_hz": 17.320687755108374,
+      "mean_period_ns": 5.778441590763602E7,
+      "jitter_std_ns": 5.188229110616742E9,
+      "min_period_ns": 2411407,
+      "max_period_ns": 981625813291,
+      "timestamp_regressions": 31,
+      "unreliable_accuracy_count": 0
+    },
+    "battery": {
+      "count": 2,
+      "measured_hz": 0.001607513822087096,
+      "mean_period_ns": 622078632395,
+      "jitter_std_ns": 0,
+      "min_period_ns": 622078632395,
+      "max_period_ns": 622078632395,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    },
+    "gyroscope_uncalibrated": {
+      "count": 18869,
+      "measured_hz": 9.165734460244325,
+      "mean_period_ns": 1.0910200424607801E8,
+      "jitter_std_ns": 1.0515934032083027E10,
+      "min_period_ns": 2411355,
+      "max_period_ns": 1387464175938,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    },
+    "light": {
+      "count": 135,
+      "measured_hz": 0.06482308493632259,
+      "mean_period_ns": 1.542660305325373E10,
+      "jitter_std_ns": 1.2408390582399005E11,
+      "min_period_ns": 209457865,
+      "max_period_ns": 1389880701303,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    },
+    "magnetometer_uncalibrated": {
+      "count": 4583,
+      "measured_hz": 2.225873793602375,
+      "mean_period_ns": 4.492617698605412E8,
+      "jitter_std_ns": 2.133636936273893E10,
+      "min_period_ns": 10000000,
+      "max_period_ns": 1387489864584,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    },
+    "pressure": {
+      "count": 1548,
+      "measured_hz": 0.7515814449804272,
+      "mean_period_ns": 1.3305277913374274E9,
+      "jitter_std_ns": 3.6297856225195114E10,
+      "min_period_ns": 31030938,
+      "max_period_ns": 1387482971875,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    },
+    "proximity": {
+      "count": 65,
+      "measured_hz": 0.031230206644233833,
+      "mean_period_ns": 3.20202812421875E10,
+      "jitter_std_ns": 1.7731487624787906E11,
+      "min_period_ns": 52996406,
+      "max_period_ns": 1381603992813,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    },
+    "rotation_vector": {
+      "count": 4600,
+      "measured_hz": 2.234147785081482,
+      "mean_period_ns": 4.4759796405479455E8,
+      "jitter_std_ns": 2.1298097011051964E10,
+      "min_period_ns": 1327083,
+      "max_period_ns": 1387569946719,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    },
+    "user_annotation": {
+      "count": 2,
+      "measured_hz": 4.8610518629781644E-4,
+      "mean_period_ns": 2057167930291,
+      "jitter_std_ns": 0,
+      "min_period_ns": 2057167930291,
+      "max_period_ns": 2057167930291,
+      "timestamp_regressions": 0,
+      "unreliable_accuracy_count": 0
+    }
+  },
+  "initial_calibration_statistics": {
+    "accelerometer": {
+      "count": 2138,
+      "mean": [
+        -0.7648618971879466,
+        7.802008074811957,
+        5.929200463951802
+      ],
+      "standard_deviation": [
+        0.6421546501325623,
+        1.1383099466862927,
+        1.1554819889711003
+      ]
+    }
+  }
+}

records_part_001.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c36b07deada6a4503b3d862b901f0322e172fc00771c7dd9247bd83fb7f8b16
+size 2215149

schema.json

@@ -0,0 +1,29 @@
+{
+  "schema_version": "1.0.0",
+  "format": "newline-delimited JSON compressed with GZIP",
+  "timestamp_contract": {
+    "timestamp_monotonic_ns": "Primary ordering clock from Android sensor hardware or elapsedRealtimeNanos",
+    "timestamp_epoch_ms": "Wall clock for cross-device alignment; clock changes must be considered"
+  },
+  "fields": {
+    "sequence": "int64, repository order",
+    "source_sequence": "int64, service acquisition order; -1 for UI annotations",
+    "record_type": "sensor | location | battery | annotation_auto | annotation_manual",
+    "timestamp_epoch_ms": "int64 milliseconds since Unix epoch",
+    "timestamp_monotonic_ns": "int64 nanoseconds since boot",
+    "sensor": "stable sensor key",
+    "accuracy": "Android sensor accuracy or estimated location accuracy",
+    "unit": "unit contract for values",
+    "values": "array<number|null>; source sensor values retain Float32 precision, location retains Float64",
+    "known_activity": "protocol label supplied by contributor",
+    "auto_motion": "coarse automatic estimate, not ground truth",
+    "phase": "baseline | pre_failure | failure | recovery",
+    "event_type": "nullable event label",
+    "confidence": "nullable heuristic confidence 0..1",
+    "details": "nullable explanation"
+  },
+  "integrity": {
+    "blocks.jsonl": "SHA-256 chain over canonical JSONL bytes in blocks",
+    "integrity.json": "SHA-256 and byte size for every exported artifact"
+  }
+}