As a rule of thumb, GPS works well for devices used outdoors, where they have a clear view of the sky. So for devices for predominantly outdoor use, GPS becomes an option. Typically for agricultural context, rural context and some industrial context, GPS is very appropriate tracking option.
However, because GPS is a technology that for many people is synonymous with tracking, particularly in the transport industry, they ask for or say that they need a GPS tracker without thinking about the optimal solution. For very expensive assets like trucks, that might make sense, as a GPS tracker may cost $3000. But we’re getting to the point where users wish to not just track the truck but also wish to track the container, the pallets within it and individual boxes on the pallets. For those kinds of things, a lot of the time is spent inside buildings, which can be out of GPS range. GPS would be a really poor choice technology for some applications and there are other better options. In application such as tracking moving vehicles, response time and latency become really important.
Another factors is that GPS is energy intensive and receivers typically will draw tens of milliamps and they might need 30 seconds or so to do an acquisition. By comparison LPWAN approaches typically draw in the order of 125 milliamps but only on average around about half a second. So LPWAN approaches will often use less energy that you need for GPS which opens up design options.
Edited by Tim Kannegieter