> On the other hand, bodycams run out of batteries or the electronics glitch or automated video downloads get skipped, in ways that simply don't happen with paper notebooks or data stored in central databases.
Systems are going to fail, but they need to fail safely. If a bodycam is low on battery or not acting properly, there needs to be some notification for both the operator and department to know that the bodycam failed or is about to fail, and their procedure has to handle such cases. So for example have an indicator light that comes on at 20% battery power and officers need to return to the station and swap cameras when that light comes on, and battery levels at the end of a shift are logged automatically with a notification if any are returned below 10%.
Of course genuine failures are going to happen from time to time. If an officer falls and damages their camera in the middle of a shootout, they're not going to call a timeout so they can fix it. But such instances should be rare, documented, and reviewed. The department should at any given time be able to produce a list of what footage they are missing and why they're missing it, and such reports should be audited by third parties. During the review, evidence should be collected - pictures of any damage, software logs, statements, etc. as well as actions taken during that missing time, which should be reviewed with increased scrutiny. So long as the discrepancy is properly documented, it's fine, though procedures should be updated to prevent a repeat of this particular failure mode.
Basically if a critical piece of footage is lost it should be easy to show that this was an isolated incident and that it won't happen again. If they can't, then there is a fundamental problem at the top level of management. This is all standard practice in other high stakes fields like pharmaceutical production.