Several different issues here which overlap but still are independent.
One) network protocol. The sensors communicate via zigbee, not WiFi. So adding another Wi-Fi access points won’t help them in anyway. Well actually, it could make things worse because Zigbee and 2.4 Wi-Fi can interfere with each other. The stronger the 2.4 WiFi, the harder the zigbee has to work to get its own messages through. (5.0 WiFi doesn’t interfere with zigbee, so use that for preference.)
You can improve zigbee signal switch strength to the sensor by adding another Zigbee device closer to it so that messages can be relayed.
It’s possible that if the sensor is right at the edge of your network that it is having to repeat Messages multiple times, which would use additional battery life, but you would see this in reduced functionality as well.
- Device type. The two different sensors are actually quite different device types. One is just a simple open close sensor. That’s the one listed with the longer battery life.
The other is a multi sensor. It has an open close sensor, but also a 3 axis accelerometer. Much more complicated device, actually reporting on five or six more measured values. Consequently, however you look at it, it’s going to use more juice.
battery type. Because the open/close sensor is simpler and uses less juice, it can be run off of a lithium coin battery. Very efficient energy use. The multisensor, because it needs more power, runs off of two AAAA batteries. more powerful, but less efficient.
battery reporting. SmartThings reports battery life in tiers, not 1% at a time. So typically the first time you check it, it reads 88%, because that’s the first tier under 100%. After that, it may bounce around from day to day between two tiers before settling again for awhile.
environmental factors. Batteries work best at right around 68 degrees F. Go lower and you get reduced power. Go higher and the voltage available isn’t reduced but battery life is. In home automation setups it’s really common for a sensor in the garage to have about half the battery life of ones in the interior of the home. Sensors in a window that gets full sun may be colder at night and hotter during the day than a nearby door sensor, sometime by as much as 15 degrees each way.
A lot of people don’t know this is one reason so many battery-powered sensors include a temperature reading. It’s so you can track the environmental conditions you’re exposing the batteries to. So take placement into account when evaluating battery performance as well.
There are some other factors that can be involved, but those are the first steps.
Consider the device (a multisensor is never going to have the same battery life as a simpler device). Strengthen each network (zigbee, zwave, wifi) independently. Then watch the trend, as well as the tiers. Dropping ten percent in a week might just mean you really dropped 0.5% and hit the next tier. It can even be due to temperature variation, pretty common with window sensors. Dropping 20% in two weeks, though, is at least a real drop of a full tier. And that would be significantly faster than it should be. So that’s when you start looking at range, battery quality, polling, temperature variation, even the possibility of a bad device.