Correct dataset span and add verified summary statistics

Collection actually ran 2025-04-21 to 2026-01-09 (~8.6 months), not around 5 months. Adds a summary table computed from the released files, documents temporal gaps, clarifies the logging cadence vs the device read-out rate, and notes that the high-magnitude raw channels are the base/reference channels.

README.md

@@ -16,13 +16,54 @@ size_categories:
 viewer: false
 ---
 
-# Temi VOC/Smell Datasets (2025)
+# Temi VOC/Smell Datasets
+
+This dataset includes processed smell collection metrics recorded by 2 versions of the SmartNanotubes [Smell Inspector](https://cdn.prod.website-files.com/67a0b744e71371ac7cb22fda/6810bcae6ccbea1631617e8b_SNT_Manual_Smell%20Inspector_V2.1.0.pdf). The sensor was attached to the tray of a [Temi robot](https://www.robotemi.com/product/temi/), which roamed our office on a set path. The robot patrolled roughly once an hour during working hours and recorded a few dozen smell data points on each run. At each smell data point (in the `processed/new-sensor.csv` file), the robot also took a picture using its front top camera of the area that it was in.
+
+Data collection ran from **2025-04-21 to 2026-01-09** (263 calendar days, ~8.6 months), with data recorded on 148 of those days.
 
-This dataset includes processed smell collection metrics recorded by 2 versions of the SmartNanotubes [Smell Inspector](https://cdn.prod.website-files.com/67a0b744e71371ac7cb22fda/6810bcae6ccbea1631617e8b_SNT_Manual_Smell%20Inspector_V2.1.0.pdf). The sensor was attached to the tray of a [Temi robot](https://www.robotemi.com/product/temi/), which roamed our office on a set path for around 5 months. The robot patrolled every hour and recorded a few hundred smell data points on each run. At each smell data point (in the `processed/new-sensor.csv` file), the robot also took a picture using its front top camera of the area that it was in.
 Old Sensor: Firmware version 3.0.1
-New Sensor: Firmware version 2.1.1 (we replaced the old sensor with the new sensor due to faults in the old sensor, even though it was a downgrade in firmware version)
+New Sensor: Firmware version 2.1.1 (we replaced the old sensor with the new sensor due to faults in the old sensor, even though it was a downgrade in firmware version). The two firmware numbers do not reflect a linear progression in capability.
+
 Raw and processed CSV data and code for cleaning can be found on [github](https://github.com/Kentucky-Open-Science/Temi-VOC-Datasets).
 
+## Contents
+
+- `processed/old-sensor.csv` — 64 raw channels + temperature + humidity
+- `processed/new-sensor.csv` — the above plus robot x/y position and frame filename
+- `processed/old-sensor-features.csv`, `processed/new-sensor-features.csv` — Configuration A feature mapping applied
+- `frames.zip` — 74,870 JPEG frames, one per row of `new-sensor.csv`
+
+## Dataset Summary
+
+All figures below were computed from the released files.
+
+| Property | Old sensor | New sensor | Combined |
+|---|---|---|---|
+| Collection period | 2025-04-21 to 2025-05-30 | 2025-06-02 to 2026-01-09 | 2025-04-21 to 2026-01-09 |
+| Calendar span | 39 days | 220 days | 263 days (~8.6 months) |
+| Days with data | 23 | 125 | 148 |
+| Cleaned rows | 12,360 | 74,870 | 87,230 |
+| Patrol runs | 155 | 972 | 1,127 |
+| Median rows per run | 67 | 61 | — |
+| Median logging interval | 6.0 s | 6.0 s | 6.0 s |
+| Synchronized frames | N/A | 74,870 | 74,870 |
+| Image archive size | N/A | 4.3 GB (`frames.zip`) | 4.3 GB |
+| Frame resolution | N/A | 640 × 480 | 640 × 480 |
+| Robot position | No | Yes | New sensor only |
+| Patrol footprint | N/A | ~33 m × 62 m | N/A |
+| Temperature range | 25.3–28.6 °C | 25.4–27.8 °C | 25.3–28.6 °C |
+| Humidity range | 23.3–53.7 % | 7.3–37.3 % | 7.3–53.7 % |
+| Firmware version | 3.0.1 | 2.1.1 | — |
+
+The device's own read-out cadence is 1.8 s for all 64 channels (per the manufacturer's manual), but the logging service sampled it during patrol runs rather than continuously; the median interval between consecutive logged rows within a run is 6.0 s. All rows carry timestamps between 08:00 and 17:00 local time.
+
+Every row in `new-sensor.csv` references a distinct frame in `frames.zip`, and the archive contains exactly 74,870 images — the row/frame correspondence is one-to-one.
+
+## Temporal gaps
+
+Sampling is not uniform. Data were recorded on 23 of the 39 calendar days spanned by the old-sensor split and 125 of the 220 days spanned by the new-sensor split. Within those periods, intervals longer than one hour occur 42 times (old sensor) and 217 times (new sensor); the longest single gaps are 10.0 and 12.7 days respectively. Gaps come from charging cycles, firmware updates, network outages, and periods when the robot was unavailable. Nothing is interpolated — rows are either present or absent — so gap locations are recoverable directly from the timestamp sequence. Handle them explicitly in any time-series modeling.
+
 ## Data Dictionary (`processed/old-sensor.csv`)
 | Column              | Data Type                      | Purpose                                                                        |
 |---------------------|--------------------------------|--------------------------------------------------------------------------------|
@@ -41,7 +82,17 @@ Raw and processed CSV data and code for cleaning can be found on [github](https:
 | humidity            | float                          | Measurement from the humidity sensor included in the smell sensor (%)                        |
 | robot_x_position    | float                          | X position of the robot w.r.t. its home base                                                 |
 | robot_y_position    | float                          | Y position of the robot w.r.t. its home base                                                 |
-| frame_filename      | string                         | Filename of the picture taken by the robot when a smell was measured (see frames/ directory) |
+| frame_filename      | string                         | Filename of the picture taken by the robot when a smell was measured (see frames.zip)        |
+
+## A note on raw channel magnitudes
+
+Raw channel magnitudes are strongly non-uniform. In `new-sensor.csv`, seven channels (0, 1, 14, 15, 19, 48, 49) have median values more than an order of magnitude above the typical channel, with channels 19, 48 and 49 sitting near 2.2e6 against a median channel value of ~8.1e3. In `old-sensor.csv` the spread is much smaller (largest per-channel median is 24× the smallest, and no channel exceeds 10× the typical value).
+
+In both splits these high-magnitude channels are exactly the base/reference channels that the manufacturer's Configuration A mapping marks as 999 and that the feature-preparation script drops. The spread is a property of individual channels and their assigned role, not of any one detector chip. If you work with the raw 64-channel vector instead of the mapped feature set, scale per channel before analysis.
+
+## Channel mapping
+
+Under the manufacturer's Configuration A mapping, 45 of the 64 channels carry analyte-relevant signal corresponding to 15 distinct feature IDs, each measured by 3 replicate channels; the three replicates for a given feature ID sit within a single detector type, not one per detector. The remaining 19 channels are base/reference channels excluded from analysis. The `*-features.csv` files have this mapping already applied.
 
 ## License
-This dataset is licenced under the Apache License 2.0.
+This dataset is licenced under the Apache License 2.0.