Measurement-based evaluation of Google/Apple Exposure Notification API for proximity detection in a commuter bus

Abstract

We report on the results of a measurement study carried out on a commuter bus in Dublin, Ireland using the Google/Apple Exposure Notification (GAEN) API. This API is likely to be widely used by Covid-19 contact tracing apps. Measurements were collected between 60 pairs of Android handset locations and are publicly available. We find that the attenuation level reported by the GAEN API need not increase with distance between handsets, consistent with there being a complex radio environment inside a bus caused by the metal-rich environment. Changing the people sitting in a pair of seats can cause variations of ±10dB in the attenuation level reported by the GAEN API. Applying the rule used by the Swiss Covid-19 contact tracing app to trigger an exposure notification to our bus measurements we find that no exposure notifications would have been triggered despite the fact that all pairs of handsets were within 2m of one another for at least 15 mins. Applying an alternative threshold-based exposure notification rule can somewhat improve performance to a detection rate of 5% when an exposure duration threshold of 15 minutes is used, increasing to 8% when the exposure duration threshold is reduced to 10 mins. Stratifying the data by distance between pairs of handsets indicates that there is only a weak dependence of detection rate on distance.

Fig 1

(a) Bus on which measurements were collected. (b) Relative positions of participants during tests.

Fig 2. (a) Measurements of attenuation between two handsets as the distance between them is varied along the centre aisle in the upper deck of the bus, (b) shows the setup used.

The vertical dashed lines indicate when the distance between the handsets was changed, starting at 0.5m and then increasing by 0.5m at each step. The solid horizontal lines indicate the mean attenuation level at each distance. Measurements taken using the standard Android Bluetooth LE scanner API.

Fig 3. Attenuation durations reported by GAEN API on completion of the first test on the lower deck of the bus.

Fig 4. Attenuation durations reported by GAEN API on completion of the second lower deck test (with the same participants as in the first test, but with their seating positions swapped about).

In (d) person 5 is using a Samsung Galaxy A10 rather than a Google Pixel 2.

Fig 5. Attenuation durations reported by GAEN API on completion of the first test on the upper deck of bus.

Fig 6. Attenuation durations reported by GAEN API on completion of the second upper deck test.

Participant in seat 4 is absent for this test but otherwise the participants are the same as in the first test, but with their seating positions swapped about. We have kept the x-axis labelling the same as in Fig 5 to facilitate comparison.

Fig 7. Exposure notification rate obtained when applying a range of exposure notification rules to the GAEN bus dataset.

Fig 8. Exposure notification rate broken down by distance between handsets for the GAEN bus dataset.

Threshold exposure notification rule, t2 value marked on x-axis, ES threshold 10 minutes.