Not selling cameras, is a topic that can come up in security discussions. It is a very instructive topic because it explain data volume, latency, bandwidth availability, storage, power requirements, background infrastructure and legislative constraints. Cameras are everywhere; dashcams, street cams, webcams, bodycams, phone cams, etc. The urban world appears to be one big movie scene, captured from a variety of angles. When someone thinks about “IoT security,” their first impression is visuals – but there is a vastness beyond that application.
First, you have to ask what a camera does: What does a camera do? It collects a vast amount of information in real-time. This activity requires bandwidth. Depending on your use case, that data can be transferred as a single moment (a snapshot, pic, selfie, still frame, etc.) or captured as motion footage. Motion capture is the more “security” focused tool, but there are many applications for still images. The collection, collation, segmentation and analysis requires backend organization. Companies that do video, do video. Companies that do cameras do cameras. We agree that the activity and use of the tool is very different than building, maintaining, mounting, wiring and supporting a tool.
They’re power hungry
A camera, even a small phone camera is power hungry. There is a reason that small phones back in the 90s – 00s could last a week with a single charge. A new phone lasts about a day, with vastly increased battery size. In fact, your phone is a giant battery behind a giant screen. That phone from the 00s was light, small,” agile” and almost indestructible…it is much closer to the rugged IoT model, prioritizing smaller data packet audio and asynchronous tools like text messaging (still popular).
Our business is all about remote, rugged, rural, rainy day, rough reliability – everything is about power consumption. Once we nail that down we can build your product. If I position my camera on the back of a waste collection vehicle in November, it lasts a about a day, due to the environmental constraints. You might say; Use a GoPro. How do you clean the screen? If I use a ruggedized ultrasonic sensor or weight based sensor, I’m comfortable looking at 5 years, and only a handful of service callouts across your entire fleet.
Storage costs
The second part of that dynamic is storage. Where does all that information go? (network or stored onsite) How quickly does it get sent? (latency) where do you store it? (Backup, analysis and retrieval). If you’ve got more data, you need bigger and better tools to sift through it. That is why the AI is becoming such a pivotal part of camera based security tools. It’s even writing it’s own ad copy. According to AI, you can see a kingfisher sucking nectar in real-time if you really want to…but there is an issue with that concept.
Siting – cameras are usually fixed to a particular field, even if they can pivot. That is why fancy buildings buy lots of them. If they’re mobile, they’re fixed horizontally or vertically, unless your AI is capable of twisting a joystick and finding “the perp” in real-time …which the AI says it can do, effortlessly.
Shifting contexts
The trouble with all that advertising is that the core of any security task relies on real people doing real work, in real-time, in real world scenarios. The guy who steals a fancy e-bike off your balcony knows that cops don’t waste time tracking down hoodies these days. And now you have no vehicle to give chase either. The AI isn’t chasing anyone either, unless the thief uploads a congratulatory selfie to Instagram. The AI sits within a context, but like those old games of “cops and robbers,” the context shifts. Privacy is a competing context. So too is who you can track, why and where.
In a different context, the market for camera based electronics is dense and deep. You compete with the absolute cream of global brands when you start dabbling in optics, screens and scanning tools. You have to respect the other players in that room. A small hardware team up against giants is not viable. In sensing and tracking, being agile, rugged, custom, and in-house is are real advantages. We can partner with ruggedized camera vendors and cherry pick the best camera package, if the client requests it.
When cameras can’t see?
So, what advantage do we bring? “When cameras can’t see…” is a big topic, what Derrida would call the hidden subtext. An unpowered generator in a field, the excavator once it’s left the yard, cabinet on a remote pole, a flickering forest fire, the slow puncture on a semi trailer, an asset inside a container, the leaking water pipe underground…. the list is endless, but if you’re fixated on cameras, there is nothing else you can see.
Sensing is not a limited binary. A sensor alarm or temperature shift can tell a monitoring center which camera to look at, unless you’re running the plotline of the Thomas Crown affair in your head. In that specific context, an accelerometer or a 25 cent RFID tracking tag might have saved their precious painting without any drama. The ability to use other senses informs human beings, it also works in the other domains. Synergies between sensors and division of labor are the keys to the puzzle.
