Hello - can I know what you are exactly testing. I have a lot of experience geo testing and have found that physically triggering the geo fence always works the best and provides the most accurate results. This being said, using something like a geo simulator doesn’t hurt to get a good understanding to how your app is performing.
What geo layer are you using in the app? Something like Pathsense?
I’ve done a lot of geo fence testing on a bike by traveling away from the device only using the mobile phone and app active in my pocket and past 250ft radius to arm the device (was a security device that activates when geo fence is hopped at 250ft).
I’ve also learnt that it’s very dependent on network, model of phone, phone platform i.e. iOS/android and where you are located in terms of mobile masts to help triangulate location etc.
I’ve had issues where even when I have hopped the geo fence, the app has not picked up the movement till a few minutes or sometimes even hours later - usually this is down to triangulation issues, bad signals, latency on phone or network etc.