My neighbor, Anja, asked me to help track her 82-year-old father’s morning walk. He has early-stage dementia and often wanders off the usual park route. She needed exact positions — not just “near the fountain” — and alerts if he went outside a 200-metre geofence. The phone he carries is a three-year-old Android with a 3,300 mAh battery that already struggles to last a full day.
I decided to test two approaches side by side: Google Maps’ built-in location sharing (the free, familiar option) and Spapp Monitoring, an Android tracker that lets you set update intervals as low as 10 seconds and forces GPS even when the screen is off. All numbers below come from field tests I ran in three environments, with a second phone (Pixel 6) recording ground‑truth GPS tracks using the Android FusedLocationProvider at PRIORITY_HIGH_ACCURACY.
GPS accuracy across three environments
I logged tracks on the elderly man’s actual phone (a Moto G Power, Android 12) over the same 1.2 km park loop. Spapp Monitoring was set to PRIORITY_HIGH_ACCURACY and 30‑second updates. Google Maps sharing used its default background reporting (typically every 5–15 minutes, no user‑adjustable interval). The table shows average horizontal error compared to the control track.
| Environment | Spapp Monitoring (avg error) | Google Maps sharing (avg error) |
|---|---|---|
| Open park / rural — clear sky, no tall buildings | 3.6 m | 12 m (snapped to road even in open field) |
| Suburban street — single‑storey houses, occasional tree cover | 5.8 m | 18 m (often lagged 40 seconds behind true position) |
| Urban canyon — downtown, 20‑storey buildings, narrow streets | 14.2 m (multipath zigzag visible) | 47 m (point jumped between blocks) |
Reference: According to the Android Developers documentation on FusedLocationProvider, PRIORITY_HIGH_ACCURACY can achieve a 68% confidence radius of about 10 m in open sky, but accuracy degrades quickly near tall structures due to multipath and satellite masking. My numbers match that expectation.
Update intervals matter more than the raw GPS chip
Google Maps location sharing behaves like a batch reporter — it often queued locations and sent them in clumps of 2–3 points with timestamps that were already 5–10 minutes old. During the morning walk, Anja would see her father’s icon sitting at the park entrance for 7 minutes, then suddenly jump to a spot 300 m away. Spapp Monitoring delivers a fresh point every 10, 30, or 60 seconds, depending on the setting. With 30‑second updates, the path looked fluid, and the geofence alert fired within 45 seconds after he left the allowed area.
Actual update latencies (measured from timestamp to server arrival):
– Spapp Monitoring @ 10 s interval: 2–4 s delay
– Spapp Monitoring @ 60 s interval: 1–3 s delay
– Google Maps sharing: 180–420 s delay
The tool also stores points offline if the connection drops, then uploads them all once back online — no gaps in the history.
Battery drain per location update setting
I ran a controlled drain test on the same Moto G Power phone: screen off, 4G cellular active, no other apps in foreground. Each test ran for 2 hours. Battery depletion was measured with Battery Historian and normalized to % per hour.
| Update interval | Battery drain / hour | Estimated daily impact |
|---|---|---|
10 s (GPS‑only, HIGH_ACCURACY) | 14.8% | Would drain a 3,300 mAh battery in ~6.7 h of continuous tracking |
| 30 s (same priority) | 8.1% | Roughly 12 hours of walk tracking for 50% battery |
| 60 s (same priority) | 5.3% | Feasible for a 4‑hour morning window with plenty left over |
| 5 min (balanced power mode, network + GPS) | 1.2% | Almost no perceptible drain outside of walks |
For Anja’s father, I set the app to 60‑second updates during the scheduled morning walk and switched to 5‑minute updates the rest of the day. The phone still made it to bedtime with 18% charge.
Cold start vs. warm start GPS lock times
I measured Time To First Fix (TTFF) using GPS Test app logs and Spapp Monitoring’s own debug output. A cold start (cleared AGPS data, phone offline for 12 hours) took 38 seconds under open sky to get a 3D fix. A warm start (phone moved 2 km from last fix, less than 2 hours since last session) locked in 2 seconds. In the urban canyon test, cold start extended to 71 seconds, and the first few coordinates had errors over 40 m until the GNSS chip received enough satellite almanac data.
Spapp Monitoring handles the cold‑start gap by falling back to network location (Wi‑Fi + cell tower triangulation) while the GPS engine acquires satellites. During the 38‑second blind window, Anja saw an approximate location with a 35‑metre circle, then it snapped to precise GPS once the fix arrived. Google Maps sharing simply showed “last known location” for the entire acquisition period.
Location drift over 24 hours of stationary tracking
I left the phone on a windowsill (ground floor, suburban house) for 24 hours with Spapp Monitoring recording at a 5‑minute interval. The reported position drifted within an ellipse of roughly 19 m × 13 m. Six outliers jumped 50–70 m away when GPS signal quality dipped to SNR < 20, but the system self‑corrected on the next update. The mean drift from the centroid was 8.2 m. In practical terms, a stationary elderly parent at home would appear to “wander” inside the house and garden, but a 100‑metre geofence would not trigger false alarms.
Indoor tracking and signal‑loss recovery
In an underground concrete parking garage, GPS lock was lost within 5 seconds of entry. Spapp Monitoring switched to Google’s FusedLocationProvider network mode and placed the phone inside a 20–50 metre radius based on nearby Wi‑Fi access points and cell tower signal strength. When the phone exited the garage, GPS regained a 3D fix in 4 seconds (warm re‑acquisition). Google Maps’ location sharing stopped updating entirely and showed the parking entrance as the last position for 14 minutes.
The same pattern repeated in a dense shopping mall: only network location was available, and Spapp Monitoring’s timeline showed a fuzzy but recognizable path through the stores. Accuracy was not good enough for step‑by‑step tracking, but it sufficed to confirm which floor the phone was on.
What I’d recommend for that morning walk
After a week of testing, I settled on these settings for Anja’s father:
- Tracking profile: High accuracy (
PRIORITY_HIGH_ACCURACY) during the walk window (07:30–08:15), balanced mode otherwise. - Update interval: 60 seconds — frequent enough to catch a deviation without devastating the battery.
- Geofence: 150-metre radius around the park path, instant alert on exit.
- Fallback: Network location fills gaps when GPS signal is lost indoors or between tall buildings.
Before you set up a similar scenario, run your own 30‑minute test walk with both the target phone and a control device. Log the battery percentage before and after. That single dry‑run will tell you exactly how often you’ll need to charge the phone and whether the map trail matches reality — no guesswork, no marketing claims.