Just a followup on my testing.
Installed a new battery on Jan 25th and it lasted only 12 days.
The device was placed face down so it couldn't detect any motion.
The device was placed in a stable temperature area of 65F +- 1 degree.
The device was in close range of several repeaters and within 10 foot of the hub.
Clearly there is a problem. Next step is to remove it from Hubitat onto a V2 SmartThings hub.
I only have a single repeater..my mesh is possibly unstable im not sure but it has quick enough response. 1 repeater 12 sensors 21 bulbs. My hub and repeater are upstairs. Maybe trim your repeaters by half. Putting the remaining in between all devices. Factory reset the repeaters and re-pair. Then leave your hub off for a few minutes to cause a panic and repair the mesh. This is just suggestion as I'm more familiar with samsung. Good luck! 
I have three of these. Two installed mid November '19, and one late December '19. All three currently read 99% battery. One is in a hallway that is transited at least a dozen times a day. The others, one in the laundry room and one in a home office. Both activated multiple times daily. The batteries are still factory stock. I'm using the Generic Zigbee Motion Sensor driver. One is within a 8 ft of the hub. The other two are within 8 and 15 ft of a Smart Things plug used as a repeater. Zigbee channel 20.
I am still trying to figure out why I am having so many battery problems with this specific device. I have determined that the batteries last much longer when they are close to the hub. It seems the further away they are the faster the battery drain (seems logical). But even when they are close to repeaters (centralite cube outlets) the battery life is much poorer. Sometimes lasting only three days.
I have nearly 70 zigbee devices, many are battery operated and get superb battery life. I have first and second generation SmartThings motion sensors with batteries lasting almost two years. I have put the latest generation motion sensors right next to the other ones and batteries last 7-14 days.
Right now I am doing another test. I have two other hubs. One is a V2 SmartThings hub and the other a Hubitat hub. These hubs have only one device connected to them, the latest generation SmartThings motion sensor. Tomorrow they will have been connected for 14 days. The one connected to the SmartThings hub is down to 90% and the Hubitat hub is at 87% Both sensors are within 20 foot of their respective hub. So already I am seeing much worse battery life than yours.
It is frustrating because nobody seems to have a problem with them besides me and it is equally frustrating that of all the devices I have it is these specific motions sensors that have problems.
Anyway that is all the information I have at this point. I continue to try and find out what it is and if I find the answer I will post an update.
I had an St motion that ST updated the firmware on and that ruined the battery life. It was sending constant motion, even when inside a cardboard box! Samsung replaced it for me. I really loathe these sensors, so much so I tore them all apart, and converted to AAA batteries. If you have a 3d printer or a library near by that has one, it's an easy to print part.
Funny you should mention this. I took two of the latest generation SmartThings motion sensors and moved the guts into a small project box and power it with two CR123A batteries. Batteries in one lasted just under two months and the other one is still going with a battery level reading 87%
To be fair both of these are installed outdoors but still in close range of a repeater. Even with the extra battery capacity the life is still piss poor.
Thought I would chime in here. I just moved from Wink to Hubitat and I can tell you there is something different happening when using the SmartThings multipurpose v5 sensors. On the Wink hub 2 i was using for about a year, the sensors all were hanging around 100% the entire time. No issues and they worked perfectly. I believe Wink ignored the acceleration part of the sensor. After putting all the sensors on Hubitat the battery levels are dropping like crazy. All have plummeted from over 95% to 71% over the last week. Could Hubitat be polling them differently?
I have about 6 of the Smartthings motion sensors, They are using the generic zigbee motion sensor drivers. I have had them running for about 6 months now. They are all between 75% and 87% battery life. The one which is 75% is in a very busy area (entry into the kitchen) which probably gets triggered around 50 to 100 times a day.
I have 2 Smartthings multipurpose v5 sensors, they have been in for about 6 months on the front door and about 5 months on the backdoor. The front door is at 57% and the backdoor is at 85%, I am not sure why the big difference, but the front door is mostly open so maybe it uses more battery when the contact is "open"
With zigbee I don't believe there is polling, it will kill the batteries very fast. Activating the acceleration will surely affect battery life, but not to the degree you're seeing. Sounds like a mesh issue. And trust me I certainly realize that hubitat has weaker radios than, ST and wink, but the tradeoff in functionality is so worth it to me, as the radio issue is fixable by adding a repeater.The wink and ST failures are out of our hands. Plus HE is so amazing in terms of expansion, control & speed-IMHO
I'd suggest re-pairing the sensor, and adding a repeater in the middle of the sensor and the HE. Pair the repeater first, then add the sensor. This will create a route back to the hub, probably 2, and then the sensor won't continually repeat the messages because they weren't acknowledged. Repeated unacknowledged messages will kill the sensor battery.
Just a little update on the poor battery life I experience using the latest generation of the SmartThings motion sensor. I have take two of these motion sensors and moved one to a SmartThings hub and the other to my second Hubitat hub. The only devices on those hubs are the motion sensors, no repeaters or anything else. They are placed in a moderate traffic area getting triggered about 20 times a day. New batteries installed and test began on April 1st.
Almost two months later the one sensor on SmartThings is reading 99% battery and the one on Hubitat is at 87%
These same sensors were originally on my main hubitat hub with at least ten centralight cube outlets (repeaters). Both sensors batteries lasted between 7-14 days. Maybe a month if I was extremely lucky.
My test is not complete because the battery levels are still good even at 2 months.
So far this tells me it isn't a problem with the sensor(s) and its not a problem with the hubs but has more to do with the network they are in. Question is, why are they sensors eating batteries for breakfast on a hub with other zigbee devices and plenty of repeaters? Why is the battery life so poor when I have at least sixty zigbee devices (including 5 other older generation SmartThings motion sensors) that have battery life over a year?
Still don't have a good answer but have at least eliminated the hardware as the problem.
i just installed 4 ST door/motion/temp sensors, the 2 on regular doors no problems, the 2 inside freezer doors to monitor door open and temp are less than 50% in 24 hr.
could temp changes be the problem? lots of fluctuation in freezers, 0-28 degrees depending on door open or auto defrost cycles.
Bill
I installed one of those sensors on my garage door back in February (winter) and found that the battery level dropped to 50% within a day but has remained there for four months now,
I have never found battery level reports for any device being even close to being accurate. I have one door sensor that has been reading 0% battery for two months and it still works. I have other sensors report battery level at 78% but the device stops working. When I measure the voltage of the battery with a multi-meter sometimes they have less than a 1/2 volt.
What I do now is use Device Watchdog and report on the lack of device activity.
I also keep a spreadsheet on when I change batteries for each device. After a period of time I can know about when I will need to replace the battery based on its track record.
thanks for the tips and info. I'll monitor activity as you suggested and change when required.
Bill
Low temperature will definitely shorten the battery life.
It's too bad SmartThings doesn't make the original multisensor anymore because I have one in my freezer that has been on the same batteries since October 2018. I don't know why they work so well but it was the best sensor they ever made. Yes, cold effects battery life but some of these later generation sensors being sold could be much better than what they are.
I would say that your fridge may be a little thick for it. I doubt its the temp as I have a newer gen sensor in my freezer. constantly losing/searching for the mesh through the fridge may waste battery. a check would be to leave the sensor outside your fridge for a while an see if the battery life changes. I would also make sure your batteries arent suspect.
My guess is your problem is the Centralite repeaters. I know that they work well for many people, but I've seen a few posts about problems with zigbee 3.0 devices and Centralite outlets. I believe end devices have a 'high power" mode for joining the network and may remain in this mode if the signal is weak. If you look at your [yourhubIPaddress]/hub/zigbee/getChildAndRouteInfo, do the sensors appear as child devices?
None of my SmartThings motion sensors show as child end devices on my production hub.
I do believe you are correct that these motion sensors can change their power output.
I have been testing two of these motion sensors. I have one paired to a SmartThings hub and the another to a Hubitat Hub. Both of these hubs only have one device paired each, and that are the motion sensors I'm testing. There are no repeaters and both are end devices. I started this test back in March 2020 and both motion sensors continue to function with battery levels of >= 80%
These same two motion sensors were on my production hub and batteries lasted from 7 days to a month if I was lucky. So whatever the problem is has something to do with my main production hub which has plenty of repeaters and otherwise functions great. I have lots of zigbee devices and have earlier generation SmartThings motion sensors and have no problems with battery life on any devices except this specific latest generation of SmartThings. I don't know if these motion sensors are 3.0 devices but if they are then that could explain why they don't work well on my production system with lots of repeaters.
Yep! That is the motion sensor I have a problem with. So do you think the solution is to buy the 3.0 zigbee outlet? Could I drop a couple of those in and still leave my many other outlets in place? If so I could work with that. If not then that would be too costly to replace a dozen centralite outlet just so I can use these motion sensors.
BTW I make use of the buttons, water leak sensor and the multi purpose sensor and the battery life for those are excellent. Its just this specific brand motion sensors I have a problem with
Download the Hubitat app
