Yes, Virginia, It is possible to store 100% of police body cameras

It’s a fair question, but one with a very clesOar answer. Could we really keep all body camera footage from all officers in the cloud? You bet your beaten bottom, Virginia.

So how many officers we talking?

At the state, local and federal level, there are fewer than 1.3 million full-time and full-time equivalent officers in the US. Tracking down exact data is difficult, but it seems only about 60% of them are actually interacting with the public.

Unless the officer is interacting with the public, there’s little reason to have a camera running. Not “no reason”, mind you, but it’s not quite as important.

60% of 1.3million is 780,000 officers on patrol, but to further prove the point, I’ll use 100% of officers for the remainder of this thought exercise.

How many hours per officer we talking?

We don’t need 100% of clock-in to clock-out time. We don’t need to see bathroom or lunch breaks. We don’t need to see them filling out paperwork at the station. We don’t need a camera copy of time spent in court testifying which can often take up half a day or more. We don’t even realistically need to see them simply driving down the road.

But let’s say it’s hard to keep it all sorted out and an unlikely 90% of the 480-minutes per day ends up on camera. That’s 432-minutes per day. Five days a week would be 36-hours of footage.

A full-time equivalent officer should be working 50-weeks a year, less holidays, so let’s say 52-weeks just to cover overtime and holidays. That’s 1,872 hours we need to record… per officer… per year… times 780k to 1.3million officers.

Starting to sound a bit much, isn’t it? Don’t worry, it’s not.

How much hourly data we talking?

Cameras have tiny, efficient processors, and as a result, they make large, bulky files. The key to keeping these videos is recoding them into something more storage-friendly, like H.264, which is what’s used by iTunes, YouTube and Adobe Flash.

It requires processor time to convert these videos into more reasonable file sizes, but a data center would have the time and processing power needed.

YouTube converts 100 hours a minute into H.264, but this would require 4,630 hours per minute. Daunting, but not impossible by any means.

So the quality, wouldn’t it just suck?

Take for instance this two hour, 720p video. It’s only a one GB file on YouTube, even even at full-screen, it is very crisp and clear.

This is what all the major players use for their digital compression and very few complain that Apple doesn’t provide sufficient quality videos in on iTunes.

Could the data center even process that much video?

Almost certainly. YouTube’s servers are bogged down with a number of other tasks that simply wouldn’t apply to a bodycam repository.

  • YouTube serves videos out to over a billion visitors per month. This datacenter would serve fewer by orders of magnitude.
  • YouTube converts videos into multiple file-sizes ranging from 4k all the way down to 360. This would only need to make a single conversion copy.
  • The files coming in for processing would not be of a thousand varieties, but a small handful of pre-established formats
  • YouTube pre-checks all videos to see if color, anti-shake or other enhancements are needed, then employs them. This would not be required.
  • YouTube employs copyright-checking on every video and the audio, which wouldn’t apply.
  • YouTube dedicates huge server resources to hosting and displaying advertisements, as well as tracking them.
  • YouTube tracks all views and visitor metrics for over a billion visitors per month. Very little analytics would be required, and even less considering views would be in the millions at most, rather than billions at least.
  • YouTube makes everything available in real time, or as close to it as they can. These conversions and data requests could be scheduled for off-peak processing hours.
  • YouTube retains data essentially forever. This information would only need to be kept for months or a year, assuming it wasn’t flagged for archive due to an ongoing investigation or inquiry.
  • GN-body-cames-wide-lapel

    So you’re saying storage is cheap?

    Storage itself is cheap, though a data center is not. An internal hard drive from Amazon can get you 3tb drives for $101.72. At 655.2Gb per officer, that’s only $22.22 per officer.

    So yes, storage is cheap, but that’s hardly the complete picture.

    When Google builds a data center, they pay about $600 million to get it up and running. The previously mentioned data storage costs would be included in this figure.

    Electricity to a mega-massive data center could be as high as $40 million per year, though that would be for a vastly larger datacenter. Still, let’s give it a crunch.

    Considering we’re talking about 2.4 billion hours of footage, that’s only 1.64 cents per hour.

    Would you be okay with the officer you’re talking with having a body camera knowing it’s costing you 1.64 cents per hour? For just 12-cents a day, the datacenter hosting that footage can be preserved. If you think that’s not worth it, you might want to reserve judgment until you’ve read the costs of not having it listed below.

    But no datacenter can actually hold this much data, right?

    Wrong. Not just wrong, but wrong by miles and miles. The total data in this equation is only .85 exobytes per year of storage. The NSA’s Utah Datacenter (which we used for power consumption figures above,) holds between 12-exobytes and 1-yottabyte.

    For scale:

  • Twelve Exobytes could hold all officer bodycam footage covering a 14-year period.
  • One Yottabyte could hold all footage gathered from these 1.3 million officers spanning 738,784 years without having to delete footage going back an ice-age.
  • The storage, we got that covered.


    Yeah, but that’s the NSA. Nobody else has that much storage

    Google has around 15 exobytes of storage and they’re not even a government agency.

    For what the $1.7 trillion spent in Iraq we could have built 2,000 of these datacenters with $25 million per year left over to run them for the next ten years… and that’s to say nothing of the veteran costs we’ve incurred, which will easily be another trillion.

    Yeah, but there’s obviously other costs to consider…

    Every war has its costs, soldier. The upfront cost would be positively massive. Even at a mere $1,000 per officer (you can’t just use GoPros, you need something purpose built,) you’re talking about $1.3 billion just for the cameras.

    Sure, that’s only 4-months of cost in Iraq, but it’s still a hefty chunk of change. But let’s keep it in perspective. Last year, Social Security paid out $1.3 trillion, so based on that budget, it would only take eight hours and 45 minutes to pay for all the body cameras.

    But there are costs that can be spared:

  • Many departments already have cameras in place. No need to buy them any until they wear out.
  • Many departments already have entire support systems and storage in place, which could be reduced or eliminated.
  • New York City alone paid out almost half a billion in civil damages over a five-year period, and that’s to say nothing of the 97.3% of Americans who do not live in the big apple.
  • Use of force and officer complaints fall by 59 and 88% once there were cameras. This is due to everyone being on their best behavior and knowing it’s all being recorded.
  • Officers will be less likely to dilly-dally on the clock, leading to improved productivity.
  • This is going to tax the officers time pretty heavily though, right?

    It doesn’t have to. The departments that have already made them standard have not shown any additional burden from their use. If we roll out a national standard, there’s no reason they can’t be as effortless as putting them on a connected charging/docking station at the end of the shift.

    No uploads to worry about, it all just happens while its charging. Don’t even need an extra guy around the station to handle all the data. The IT guy can set them up the first time and replace them as they die, but in terms of daily use, a docking station can handle all of it.


    What about priiiiivacy?

    Yes, yes, what about the faces of confidential informants, children and innocent victims captured on the cameras? If Google Maps can blur every single face on the planet with no perceivable effort, these can be handled in the same way.

    Data can be compressed and stored, then only processed to blur faces once the footage is needed for an investigation, trial, or public release. The technology exists, there’s no reason to pretend it doesn’t.

    Yeah, but what would it really cost?

    Jobs, most likely. 99% of officers are good guys who appreciate their jobs and enjoy helping the public good. Unfortunately, they get lumped in with the 1% who are jackbooted jackasses who eagerly punch cuffed suspects in the face or throw grenades at those peaceably assembled.

    Officers protect their brothers because currently, for the most part, it’s his word against the suspect’s, and often that opposing word is from a career criminal. Body cameras will cost some officers their jobs, but those are exactly the guys who need to go. They are the ones destroying public trust. They are the bad apples.

    This will not happen

    It is too cheap, too easy and way too reasonable. Things are going to have to get much, much worse before they can get any better at all.

    I welcome your comments below on why this is a terrible idea, sheer brilliance, or any logistical issues in between.

    Author: Brian White

    Brian first began peddling his humorous wares with a series of Xerox printed books in fifth grade. Since then he's published over two thousand satire and humor articles, as well as eight stage plays, a 13-episode cable sitcom and three (terrible) screenplays. He is a freelance writer by trade and an expert in the field of viral entertainment marketing. He is the author of many of the biggest hoaxes of recent years, a shameful accomplishment in which he takes exceptional pride.

    1 thought on “Yes, Virginia, It is possible to store 100% of police body cameras

    1. Face it. Cameras on police officers is going to yield only jumbled images facing in the wrong direction wherever anything incriminating happens. Epic fail. But then again we’re used to that from the criminal justice system aren’t we?

    Comments are closed.