Skip to content

Tailored solutions

Tailored solutions. Start with what you want to know.

This is the TAILORED part of Infinity Neural: detection studied and tuned for your particular scene, not a system installed the same way everywhere. It is organised into six families, split by the kind of thing that has to be understood in front of the camera.

What this is, and what it is not.

Nothing standard gets installed here. We look at your camera — where it is, what angle it has, what the light is like at six in the evening and at six in the morning, what can be seen from there — and we tune the detection for THAT site. That is why it is the most accurate thing you can have in a specific case: because it is made for that case and not for every case.

What it asks in return is a connection: the video is processed off site. If your installation cannot depend on a line, or if what you need is to cover forty cameras and many cases at once instead of tuning one, IRIS Neural is probably what suits you. We do not say that to get out of work: tuning forty things one by one does not pay off, for you or for us.

Seven areas. Six families.

An area is a business problem: 'I want to know how long my customer waits', 'I want nobody coming in without a hard hat', 'I want to know why the line stops'. A family is what solves it.

They do not map one to one. Some areas share a family — people safety and intruder detection are both covered by Black — and there are installations where two or three are needed at once on the same cameras.

An open-air car park at midday, seen from a high floor. At the back, the rows of spaces with trees between them and the sea behind; on the right, two reserved bays painted blue with the wheelchair symbol. In the foreground, a raised barrier next to a control hut lets a blue car through while another waits behind and several pedestrians cross at the zebra crossing.

The six families.

If you already know which one you need, go straight in. Grey does not have its own page yet: the way in is Neural Transit, the product that counts vehicles.

Customer experience.

Everyone has an idea of how many people come into their premises. Almost nobody has the number, and those who do cannot always explain where it comes from. None of that matters until the day you have to decide how many tills to open, negotiate a rent based on footfall or justify an occupancy figure to a third party.

What gets measured here is what people do without noticing: how many come in and through which door, where they walk, where they stop, how long they wait and which areas nobody sets foot in. With the same video that is already being recorded and without fitting a single sensor.

And one thing that matters more than it looks: the number comes out the same every night. It does not depend on who was on shift, so you can compare it with last month's and with last year's. The counting is certified by the Centro Español de Metrología.

A shopping square in full sun, seen from a high balcony. On the left, the street with shops, a city bus and a pedestrian crossing; on the right, terraces with parasols and people sitting. In the centre, a fountain with jets and benches around it, and dozens of people crossing with bags. At the back, in front of the shopping centre's glass entrance, a long queue has formed between posts and tape.

People safety.

Hardly any workplace accident is a surprise. Before it there was someone without a hard hat, a forklift crossing where people walk, or a person too long in an area where they should not be. Things you can see, if somebody is watching.

The rule is set once — which area, what kit is needed, what happens if it is missing — and from then on the camera watches all the time. Anything that complies produces nothing: if every properly equipped person set off an alert, nobody would look at the alerts.

And it arrives while it is happening, not in the investigation afterwards. That is the difference between a recording for the incident report and an alert that stops it.

The inside of a manufacturing workshop seen from a high corner, with the walls opened up. On the floor there is a green-painted walkway with zebra markings at the crossings, and several operators in hard hats and high-visibility vests walking along it. At the crossing in the centre, a forklift with its beacon on stops while a worker finishes crossing. On the right, two people follow the floor from a glazed office.

Production efficiency.

The PLC knows whether the machine is on. It does not know whether the product is coming out right, whether a bottle has fallen over, whether there is a gap on the belt or whether someone has spent ten minutes repositioning boxes by hand.

Fitting a sensor for each of those things means stopping the line, cabling it and maintaining it — and each sensor measures one thing and one thing only. The camera already above the belt is seeing all of it, and it keeps the video too: when the number drops you can go and look at what was happening in front of it at that moment.

That is what a counter does not give you: the why.

A production unit seen from a high walkway. On the left, behind glass, the plant office; in the centre, two parallel assembly lines with yellow robot arms and operators in hard hats on both sides placing boxes. On the right, a roller conveyor and four people standing by a machine, looking at something on a tablet. At the back, stacked pallets and a forklift.

Logistics efficiency.

The warehouse system knows what should be there. What is really there you find out by going down to look, and that is why stock is counted twice a year and not twice a day.

The loading bay is the same: you know which lorries came, not how long each one stood there or which one got stuck. That figure lives in the bay supervisor's head, and only until Thursday.

By measuring four moments per operation — arrival, start, finish and departure — a feeling turns into a series you can compare week by week. And a gap on the shelf stops being found only when somebody walks past it.

A logistics centre seen from the air at midday, with the sea and some port cranes at the back. Filling the whole width, the loading bay frontage with numbered doors and lorries docked at nearly all of them. In the yard, two forklifts move pallets and a red lorry manoeuvres towards its slot. In the foreground, the entry hut with the barrier down and three lorries waiting in the street.

Infrastructure integrity.

A small drip, a mesh working loose, a pressure gauge needle dropping two points a month. None of those things sets off an alarm, because no system is programmed to notice that something is changing slowly.

The engineer's round happens every fortnight and covers what there is time to cover. What happens between one round and the next nobody sees, and that is where nearly every expensive fault starts.

Because the camera is fixed, today's image and the one from a month ago are the same field of view. That lets you see what the eye does not: that the stain has grown, that the needle is lower, that the anchor point has moved. The round still happens; what changes is that it arrives knowing where to look.

A utilities plot by the coast, seen from the air at midday. On the left, an electrical substation with its pylons, transformers and the high-voltage lines running off the top; on the right, a treatment works with two circular clarifiers of blue water, rectangular basins and two closed tanks. Below, the sea and the rocks lie outside the fence that runs round the whole site, with the access gate on the right.

Mobility analysis.

The classic way to count traffic is an induction loop under the surface. It works very well and it has two problems: you have to dig up the carriageway, and it counts at one point and nowhere else. So you measure where there is a loop and estimate the rest.

The camera is already on the pole. You draw on the image where each direction runs and out comes the count, split by vehicle type. And because it is an image and not a cable, the same camera can measure something else next month without going back out to the street.

It works at night and in the rain, which is when you need it and when estimating by eye is at its worst.

A city junction in full daylight, seen from a tall building. On the left, a red bus lane and a green cycle lane run down side by side next to the pavement terraces; on the right, the entrance to an underground car park with its sign. In the centre, cars wait while people cross at the zebra crossings and a yellow bus pulls into the reserved lane.

Intruder detection.

A perimeter alarm says something crossed. It does not say whether it was a person, a wild boar or a plastic bag in the wind, and that is why at three in the morning nearly every alarm is false and ends up ignored.

Here you can tell what came in, where and at what time, and the alert arrives with the image of what set it off. Whoever is on duty decides by looking, not by guessing.

And the same camera that watches at night counts people by day. You do not need one installation per job.

An industrial site at night, seen from a hillside. Inside the lit compound there are metal tanks, an industrial unit with its loading bays open and a lorry unloading. The whole perimeter is closed off with a wall and a fence topped with razor wire, and the entrance has a hut and a barrier. In the dark outside, among scrub and rocks, one person stands pressed against the wall, looking in.

And this is what comes out on screen.

Lo que Rust Neural entiende

Estado de máquinas e instalaciones.

Rust
Mide nivel de agua
Rust
Controla el color del agua de los vertidos
Rust
Monitoriza ventiladores y aisladores de alto voltaje
Rust
Verifica el estado de la vias y de las senales ferroviarias
Rust
Monitorea estado anclajes de mallas de contencion
Rust
Marcar niveles agujas manómetro
Rust
Detecta estado luces señalización industrial
Rust
Fuga agua tuberia
Rust
Puerta de nave industrial cerrandose
Rust
Detecta flujos refrigerantes
Rust
Detecta derrames de líquidos
Rust
Mide nivel agua
Rust
Detecta colores de líquidos vertidos
Rust
Verifica ventiladores y liquido refrigerante
Rust
Detecta fugas elementos gaseosos
Rust
Detecta goteras
Rust
Detecta roturas de tuberías
Rust
Detecta correcto funcionamiento máquina con bobina
Rust
Detecta si la lubricacion de la maquinaria esta activa
Rust
Monitoriza la correcta lubricacion durante corte de tuberia
Rust
Detecta roturas de tuberías y manchas de oxido

Cada una es una grabación del sistema funcionando. Pasa el ratón por encima —o tabula hasta ella— para verla en movimiento.

Filmed, in real places.

Casos grabados

Infinity Neural optimiza la ocupación y rotación de mesas en zonas de restauración

Infinity Neural mide ocupación y colas en centros comerciales con IA en tiempo real

¿Pierdes dinero por retrasos, colas y riesgos en el embarque? Infinity Neural para puertos y ferries

Análisis de flujos y zonas de interés en tiendas y supermercados con la IA de Infinity Neural

Revoluciona tu bar o discoteca con Lemon Neural

Lemon neural: optimizando la ocupación de carriles en piscinas

Control de aparcamientos con Lemon Neural

Transforma tu restaurante con Lemon y Rust Neural

Sin rodeos: Lemon Neural descifra cada movimiento humano

Get started

You do not know which one you need.

That is normal: the families are split by what they understand, not by sector. Tell us what you would want to know from a particular camera and we will tell you which one does it — or if none of them does.