The new Reolink projector camera uses sensors and AI to detect where it cannot see

Reolink has deployed a new smart home security camera at IFA 2025 which, depending on the company, can see beyond its double camera targets. This is called the projection wifi trackflex, and it looks a bit like WiFi reactivity with elite projection, a camera that I recently examined, but with a ball-shaped camera box that turns to see things that three sensors above them have detected.

This approach based on a sensor gives the camera a 270 degree degree range at any time, according to a press release that Reolink sent an email to Gizmodo. The three motion sensors are wrapped around the front and sides of the unit. A representative with whom I spoke on the IFA stand in Reolink suggested installing the camera at the corner of your house, monitoring an alley that extends from the street to a more back garage, an impossible scene for a fixed sight camera to cover, then the trackflex could detect that a car enters the aisle, swivel to face it, then watch it lead to the garage.

Like the other Reolink cameras, it stores locally recordings, either with microSD cards (up to 512 GB), or Reolink NVR and Home Hub devices. Recording videos on a storage device (NAS) attached to the via FTP network is also an option. It’s Reolink, so expect to be attacked with options in the application of the company.

As for the two trackflex cameras, they do not record a large field of vision such as the elite light wifi. Instead, one is a standard wide view while the other is a 6x zoom, capable of capturing much more zomed details than the wide goal. The two projectors looked exactly like those of the elite and can articulate to point or down, or bathe a wider area in the light. They also offer the same brightness and temperature adjustment as the elite.

The camera uses the new local AI system of the company called Reoneura Core, which allows the trackflex to make the same types of video research in natural language that we see with many other AI systems with connected camera. (See the new AI center of Switchbot.) So if you want to search for your locally stored video for a while, like a person with a brown shirt entered your garage last, you can do it. The representative of Reolink with whom I spoke took me to a pair of displays that showed me what was happening behind the scenes.

On the left, it was a view of what a user could see when using the event notifications screen; You see the live flow of the camera. While on the right, a list of events with descriptions that were mostly precise – there is were people interact during an agreement and others were Walk in the background, but he was wrong. We were not in an outdoor event, for example (although it is quite brilliant in there).

Meanwhile, on the right, the screen showed individual characteristics of things that the model identified. Seeing everyone broken down by his attributes had an atmosphere of surveillance of the police station. Reoneuro identified a person as a medium -in -law man with a green shirt with short sleeves, as well as a hat and a bag. In the lower panel, another man of average age – although this person definitively looks younger than me, and I am always at my peak, I swear – is identified as wearing a long sleeve shirt with pants, both blue and as having short hair. All these details are indeed keywords for your research later, and they all seem to be things you type if you are looking for specific events that you know the recorded trackflex.

But Woof, it’s more than a little annoying to see this in action, and it was a bit out of the key to having a Reolink representative showing it so proudly. It’s great that everything is on the device, because I prefer that it happens in a cloud server on which I have no control. It is practical and there is no doubt that I want the convenience of finding my images with casualness, but that has not stopped seeing how the sausage is made by giving me the willes.