Plant questions, engineering answers

Yes, and that is how we work. We install an industrial gateway that records what the machine produces and sends it to the platform without touching the control system. In our reference installation — a packaging machine at a food plant in Colombia — the gateway counts every bag without the machine control being modified.

The machine ends up counting every unit it produces, and reading it interrupts nothing: the data is available live while the machine keeps working. The installation is scheduled with the plant so it does not interfere with production.

It is the device that connects to the machine, records what it produces and sends it over the internet to a platform where you can see it. It is the piece that turns an isolated machine into a connected one, without replacing its control system.

The accumulated count is not lost: the equipment keeps it and picks up again when the connection returns. And the dashboard tells a transmission gap apart from a real machine stop, so hours without power do not pollute your indicators.

From a phone or a computer, with 24/7 access from anywhere. You see units per minute, the monthly total and live machine status.

Yes. Communication between the equipment on the plant floor and the platform travels encrypted, from the plant to the panel.

No. The equipment supports remote updates, so adjustments and improvements are applied without travelling to the machine.

Logbooks are filled out and approved on a web portal, and every stop is logged automatically with its duration. That replaces the handwritten record at the end of the shift, which is where short stops get lost.

Preventive maintenance acts on a fixed calendar. Predictive maintenance uses the data the machine keeps producing to anticipate a failure before it happens, so the intervention happens when the data justifies it rather than when the schedule says so.

Stops are detected automatically from the machine: nobody has to write them down. The operator adds the reason from their phone —the catalogue of causes is tailored to the plant— and very short stops are counted separately as micro-stops. The result separates what depends on the operator, what the company ordered, and what was a machine failure or a power cut.

Yes. The dashboard compares accumulated real production against the target and projects the month-end close based on the current pace and the working days left, using the Colombian calendar and its holidays. The monthly chart only shows validated production, not estimates.

Real speed, productive time and losses by cause are calculated — the components of OEE. What matters is not the acronym but the separation: how long the machine was on, how long it was actually producing, and where the remaining minutes went, cause by cause.

We also provide the hardware. The data is born in an industrial gateway that we install and answer for: if the data stops coming, that is our problem, not an empty field somebody forgot to type. A software-only platform depends on someone else capturing the data; here the capture is automatic and continuous.

With a web portal built around the format your plant already uses. In our reference installation, operators fill out the shift logbook from their phones, the supervisor reviews and approves it under her name, and submissions and approvals are notified by email. The paper is replaced, not duplicated.

Yes, and that is the difference with a digital spreadsheet. The portal crosses what the person records with what the gateway measured: stops are detected automatically from the machine and the person only adds the reason. Manual entry is reserved for what the machine cannot know.

No. Permissions are role-based: operators record, supervisors approve or return with a comment, and management consults the dashboard with the results. Every approval is logged with the name of the person who made it.

The reference system generates a monthly PDF report from the dashboard itself, with the month’s production and the day-by-day detail of the stops.

The reference installation includes an AI advisor inside the dashboard: it proposes a concrete action plan from the month’s indicators, already measured and calculated. It is not fed raw telemetry and is not asked to estimate anything on its own.

Yes, and that is exactly the job. Every manufacturer brings its own system and its own screen; we take the data from all of them into a single platform, so the plant can be seen as a whole in one place.

No. The machine keeps running on its own control and its own screen. What we do is read what it produces and bring it to the shared view, without touching how it works.

That is the most common case. They are solved with a gateway that connects them and gets them delivering data, without replacing the machine's original control.

Yes. The platform groups by plant and by line, so you can look at each one separately or all together.

The approach is non-intrusive: we measure without intervening in your machine's control. Installation is scheduled with the plant and tested on site.

For quality inspection, no. A surveillance camera does not have the optics, the shutter speed or the lighting needed to see a defect on a moving belt. For quality we install an industrial camera with its own lighting, mounted over the line and matched to your product. Industrial safety and video analytics are different: those do run on the IP cameras already in the plant.

The system inspects every product at conveyor speed and recognizes faults such as crooked labels, badly sealed containers or defective parts, without depending on someone watching.

Yes. PPE detection checks helmet, vest and glasses on everyone entering the plant and alerts the moment someone walks in without their protective gear.

Yes. You mark the hazardous zones — presses, forklifts, substations — and the system alerts as soon as someone crosses the line. It also detects a person on the floor or motionless longer than usual, even when nobody is around to see it.

Thirteen in total, across three areas: safety and EHS (PPE compliance, restricted area, person down, smoke and fire), security and access (weapon detection, perimeter surveillance, face recognition, license plate reading, ATMs and sensitive spaces, fights and aggression, pet waste detection) and operations and quality (quality inspection, people counting).

It raises the alert while it happens, not afterwards. In security scenarios it also saves the photo and video of the moment as evidence.

The camera does it on its own: it verifies helmet, vest and glasses on every person entering the plant and alerts instantly when someone goes through without their protective equipment. The check is continuous, not by sampling.

Yes. Risk areas are marked and the system alerts as soon as someone crosses the boundary. It also detects a person who has fallen or stayed still longer than normal, even with nobody nearby to see it.

It raises the alert while it is happening and saves the photo and video of the moment as evidence — useful both to react at the time and for the investigation afterwards.

No. The system is installed on the IP cameras already in the plant. If the fleet is analogue it has to be migrated to IP: the analytics needs the digital stream from the camera.

Industrial safety is about the rules being followed: PPE, restricted areas, person down. This is everything else the camera can see and count — how many people passed through an area, which vehicles came in, when something specific happened.

Yes. The camera counts people by area and by time slot, which helps both to size shifts and to understand the flow inside the plant.

Yes. Entry points and the perimeter can be watched, and vehicle entries and exits recorded.

That is one of the most appreciated uses: instead of reviewing the whole video, you search by type of event and the system takes you straight to the moments where it happened.

It recognises smoke or flame in the image from the first seconds, without waiting for the heat to reach a ceiling detector. It is complementary early detection, not a replacement for the fire system.

No. The analytics is installed on the IP cameras already in the plant. If the fleet is analogue it has to be migrated to IP: without a digital stream there is no analytics.

No. The gateway records what the machine produces without touching its control, and the machine keeps operating exactly as it did before the installation.

Only your company. Each client has a private dashboard with its own users. The demo dashboard we show on this site is a simulation with sample data.

Over an encrypted channel, from the plant to the platform.

Yes. Logbooks are filled out and approved on a dedicated web portal, and downtime logging no longer depends on someone writing it down.

The units the machine would have produced in that time and did not. That is why the calculation starts from two figures you already know: how many units your machine puts out per hour and how much each unit leaves you.

Monthly loss = units per hour × profit per unit × hours of downtime per day × production days per month. The annual figure is that same number multiplied by 12. And what shows up as recoverable is the monthly loss multiplied by the reduction percentage you pick on the last slider.

Because the calculation multiplies by the production days in the month. A stop that looks minor within a shift —half an hour, forty minutes— becomes dozens of hours a month once it repeats every day. It is the same effect that makes short stops invisible when they are written down by hand at the end of the shift.

It leaves out wasted raw material, rework, overtime and missed deliveries. It only takes the production that did not happen, which is the part you can calculate with figures you already have at hand.