Skip to content

WestPoint · Cameras · Mobotix

Mobotix does not send the video anywhere until it is needed.

German and with an approach of its own: the intelligence and the recording live inside the camera, not on a server. That changes how much network is used, what is needed behind it and what happens when the line goes down.

What kind of manufacturer it is.

The camera is not an eye sending images to a computer: it carries its own processor, it decides for itself what is relevant and it can record to its own card or to a network drive. There is no need for a recording server behind it, and the management software is not paid for per camera.

The second thing that defines it is ruggedness. Sealed units meant for hard weather and for wide temperature ranges, with very few moving parts —which are the ones that break— and no fan. In a plant with dust, pressure washing or cold, that matters more than any catalogue figure.

What it sacrifices is the price per camera and scale. It is dear per unit, the catalogue is short, and its way of working is at its best in medium, spread-out installations, not in a control room with a hundred cameras and three operators searching at once.

What being decentralised means.

That the camera decides. It looks at its scene, applies what has been configured on it and only sends something when there is something to send. In an installation with a poor line or with many small sites —petrol stations, treatment works, huts, a chain's premises— that is the difference between being able to do it and not: there is no need to push continuous video up a connection that cannot take it.

And because the video can stay inside the unit, the installation carries on recording when the network goes down. In exchange a question appears that has to be answered in writing: what happens if somebody takes the camera away with the recording inside. The answer is usually a copy on a network drive, and that is design, not a setting.

Where it behaves well.

In spread-out installations with little network and nobody in front of a screen: many small locations, each with two or three cameras, that have to raise the alert when something happens and keep their own footage while nobody asks.

And indoors where one ceiling camera that sees everything saves three: a small shop, a room, an entrance, a lift. This house's hemispherics cover a whole room from a single point and with no parts that turn, which is what breaks down most on a motorised dome.

Where we would not put it.

When detail has to be seen far away. A ceiling camera covering a whole room shares its pixels over the whole room: near the centre people are recognisable and in the far corner they are not. To identify somebody at a door fifteen metres away you need something else, and no adjustment fixes it.

And in large installations with a control room, operators searching every day and a need to integrate with the rest of the systems. It works with third party video software, but then you pay for a camera whose character is its way of working and use it as an ordinary camera, which is the worst deal possible.

The analytics it carries inside, and where they stop.

Here the on-board analytics are not an extra: they are the reason the unit exists. It detects movement with judgement, alerts on intrusion into a zone, counts crossings, and it can set off an alert, a siren or a call by itself with nothing switched on behind it.

The limit has a general part —they are drawn rules and the camera only sees its own field of view— and one of its own: the computing power of a unit with no fan and powered over the network cable is what it is, so heavy analytics do not live there.

What comes now is a recommendation with an interest of our own, so it is declared: in the same house there is IRIS Neural, the NVMS from Infinity Neural, the other company in the group. It is our own product. We would rather you knew it by reading this line.

What it brings is what will not fit into a rule, and putting what several sites see in one place. They sit together well, because the camera carries on doing its work at the edge. And if a zone alert and a local recording solve it, nothing else is needed.

What IRIS Neural is

What we do.

  • Deciding whether this approach fits, which comes first and saves the most money. If there is a good network and a control room, there is almost always a cheaper way of doing the same thing.
  • The choice of camera point by point and, on the ceiling ones, the height: it is what decides how far from the centre somebody is still recognisable.
  • Where the recording is kept and what happens if the unit is taken away. A copy on a network drive, real retention days and a recovery test done in front of the customer.
  • The setting up of the detection on site and with the scene running, because a camera that raises the alert on its own and raises it badly is a camera somebody is going to silence.
  • The maintenance: firmware, the customer's passwords, the state of the memory cards —they wear out, and nobody looks at them— and a check of the field of view.

The angle and the height are not decided at fitting time.

With a ceiling camera that sees everything, the height is the whole project. At three metres it covers the room and people are recognisable; at six it covers more room and nobody is recognisable any more. And the place matters as much as the height: the geometric centre looks nice, but if what has to be seen is who comes in through the door, the camera goes near the door.

The angle, the height and the light decide for good what will be possible with that image, and they are decided in the project with the ceiling measured, not on the day of fitting with the drill in your hand.

What are you going to ask of that image

Where to go next.

Before choosing an architecture you have to know what is asked of each image and who is going to look at it. That is what the area talks about.

Get started

Do you have many small locations and a poor line?

That is the scenario where this approach wins. Tell us how many places there are, how many cameras at each one, what connection there is and who attends to the alerts, and we will tell you whether it pays off here or whether an ordinary centralised system comes out better.