Vision AI · Webinar

Camera-based quality control an operator trains – not a data scientist

In this webinar we show DBR77 Vision AI in action: defect detection an operator trains with two buttons, asset tracking without tags, and proximity zones around people – including what a pilot installation looks like.

Webinar – 45-minute presentation + 15-minute Q&A.

  • Thursday, October 22, 2026
  • 10:00 AM ET · 4:00 PM CEST
  • Free · Online

Next session starts in

00days
00hrs
00min
00sec
Save my spot All webinars

What you'll take away

  1. 1

    How an operator trains a quality-control model with two buttons – no code, no data scientist.

  2. 2

    How asset position tracking works on the floor without RFID/BLE tags – from camera images alone.

  3. 3

    How proximity zones around people work, and how alerts reach the fleet before an incident.

  4. 4

    How camera data becomes events and coordinates in your system – with no footage leaving the plant.

  5. 5

    What a pilot installation looks like: number of cameras, edge box, PoE switch.

  6. 6

    How we make the GDPR conversation easier before it's even raised.

Agenda (60 minutes: 45-minute live presentation + 15-minute live Q&A)

  1. 0–5 min

    The problem

    Where data runs out on the floor – manual assembly, packaging, welding, older presses – and why sensor retrofits don't pay off there.

  2. 5–18 min

    Recorded demo

    Real-time detection of people, vehicles and zones from our demo environment.

  3. 18–32 min

    Quality control trained by the operator

    Two buttons, OK/NOK, a model ready in seconds – shown in the recorded demo.

  4. 32–45 min

    Privacy and the pilot

    Synthetic data, edge processing, and what installing a pilot actually looks like: cameras, edge box, PoE switch.

  5. 45–60 min

    Live Q&A

    Paweł Mroczkowski and Paweł Dera answer your questions live.

Answered before you ask

What about privacy, GDPR, and plant data?

This question comes up early in most plant conversations – for good reason. Here's the short answer:

  • No identification of individuals – the system does not recognize faces and keeps no record of who was where.
  • Synthetic data instead of footage: zones, flows, area occupancy – no recordings leave the plant.
  • Processing happens on site, on one edge device in a cabinet – not in the cloud.
  • The model learns only from your own plant's data and is not connected to any public model.
DBR77 Vision AI detects an operator in a work zone on the shop floor

Scene: NVIDIA PhysicalAI-SmartSpaces, CC BY 4.0.

“I cannot believe how easy the setup of the Vision AI is.”

– a visitor at DBR77's booth, Factory Innovation Week, Chiba, Japan (September 2026)

What does getting started cost?

A camera for one station can cost around $50*. Retrofitting sensors on that same station can cost many times more. That lowers the barrier to measuring stations where dedicated sensors never made sense.

* Cost depends on scope, number of cameras and integration. The most expensive and time-consuming step today is calibrating the camera network.

Your experts on the webinar

Paweł Mroczkowski

Paweł Mroczkowski

Executive Director Western Europe, DBR77

An experienced entrepreneur and manager who co-founded and led several companies across manufacturing, retail, services and crafts in Poland and Germany, with deep experience in digital transformation. Holds an MBA from Hult International Business School and completed executive programs at Yale School of Management and Harvard Business School. As Executive Director Western Europe and Managing Director of DBR77 GmbH in Berlin, he leads DBR77's rollout across Western Europe, connecting the Vision AI and IoT layer with plants outside the DACH region.

Paweł Dera

Paweł Dera

Chief Technology Officer IoT, DBR77

Over 20 years in defense, reverse logistics and manufacturing – leading engineering teams, process automation and predictive systems. He holds a master's degree in electronics and telecommunications and an engineering degree in computer science, and is currently a PhD candidate researching AI in smart industrial sensors. At DBR77 he leads IoT and built the Vision AI layer himself.

Who is this webinar for?

  • Quality
  • Production
  • Maintenance
  • IT/OT in manufacturing

Frequently asked questions

Is this webinar live?

Yes – this is a live webinar. Paweł Mroczkowski and Paweł Dera will host it, and they will answer questions at the end.

When is it?

Thursday, October 22, 2026, at 10:00 AM Eastern Time (4:00 PM CEST). Register on this page and you'll get an email reminder together with your join link.

How long is it?

45 minutes of presentation and 15 minutes of questions and answers from participants – 60 minutes in total.

Can I ask the experts questions?

Yes – you can ask questions in the chat during the webinar, and Paweł Mroczkowski and Paweł Dera will answer them in the Q&A.

Will I get the recording if I can't attend?

Yes, everyone registered receives the recording by email afterward, whether or not they attended.

What about GDPR and works council questions?

Vision AI does not identify individuals, processes data on site on an edge device, and learns only from your own plant's data. We will cover the details during the webinar.

What does a pilot cost?

Cost depends on scope, number of cameras and integration. After the webinar you can talk pricing directly with our experts.

Who is this webinar for?

Quality, production, maintenance and IT/OT teams looking for data from stations that have none today – manual assembly, packaging, welding, older machines.

Is this a safety system?

No. Certified safety scanners on vehicles and machines keep working exactly the same, with or without this layer – Vision AI is a measurement and data layer, not a replacement for safety systems.

Should I prepare anything beforehand?

Nothing is required. If you'd like the experts to speak to a specific case, you can optionally have a photo of a typical defect from your line on hand – it's not required.

Does registering commit me to anything?

No. Registering simply saves your spot at the webinar. Any follow-up conversation about a pilot is arranged separately, afterward.

Register now

Register below to get your personal Zoom link and reminders.

Details in our privacy notice.

Thanks for registering!

Zoom will send your confirmation and personal join link by email.

Add to Google Calendar Download .ics

Your personal join link comes from Zoom (sender: no-reply@zoom.us). Please check your spam or promotions folder, too.

No link within 15 minutes? Email us: contact@dbr77.com

You're on the waitlist!

We'll email you as soon as a new session date is set.