We have had our prototype boards up and running since February 2015. During this time we have noticed a few problems, or let us call them features:

• Thermal drift: DWM1000 does not have a TCXO. Depending on surrounding temperature and update rate (i.e. self heating due to radio usage), the measured distance can drift 3-5 cm before thermal equilibrium has been reached. This process can be as slow as half an hour or more.
• Crosstalk: This is a problem purely due to our own way of handling multiple tags. If tags get out of sync they will cause crosstalk with a random measurement error as result. Our initial remedy was taking a median value from several measurements and hope for the best. After introducing strict time alternation between tags the problem disappeared.
  
• Reflections: DWM1000 can not be mounted in contact with a wall. The reflection will be so close to the desired LOS-signal that DWM1000 can not distinguish between the two. Reported distance will thus randomly jump between two distinct values.
  
• Bootloader: If using optiboot everything is just fine. In other cases we have encountered problems with the watchdog not barking properly.