The first and primary invariant of the system is to prevent injury, be it to drivers or pedestrians. Secondly, the system should not irritate or impair the driver. Thirdly, the system should be time-independent – it should work whether or not it is day or night. Fourth, the system should be able to be operated any number of times. Finally, the warning system should only function while the vehicle is in reverse.
As with any complex system, there will be uncertainty. Radar sensors can lack precision, and thus there will be some level of error about how far aware a detected obstacle is, or in some cases (such as with smaller or abnormally shaped obstacles), whether an obstacle is detected at all. Camera vision can be obscured by environmental factors – for example, a highly reflective surface might cause lens flare and temporary blind the camera. Furthermore, one has to contend with the users. The system needs an override in cases of emergency or for authorized repairs, but the driver is likely (and necessarily) given access to the override mechanism, and clearly the system will not function of the driver has turned it off. The driver might also just ignore the video display and warnings, no matter how intrusive they are. Finally, real driving behavior might be difficult to predict. If, for example, a driver tends to be reckless, and accelerates much more quickly and to greater speeds than average while backing up, countermeasures and warnings might be ineffective (if they are to be effective under the widest possible range of circumstances).