Easy vision‑guided pick and place with IDS NXT and urobots

AI: cameras & software

This video introduces EyeBot, a flexible vision‑guided pick and place solution developed by urobots and powered entirely by IDS NXT cameras. You will learn how EyeBot combines on‑camera AI (Artificial Intelligence), 2D object localisation and robot integration to deliver fast, precise and user‑friendly automation without the need for a PC or advanced AI expertise.

The presentation demonstrates how objects are detected, labelled, trained and deployed directly on the IDS NXT camera, and how robot developers can implement pick and place tasks using simple teaching methods. You’ll also discover how the system handles flexible objects, supports multiple models, and enables scalable automation for a wide range of industrial applications.

Topics covered

  • How EyeBot uses IDS NXT for AI‑based object localisation
  • Training, labelling and deploying AI models directly on the camera
  • Setting up pick and place tasks with intuitive robot teaching
  • Working with flexible objects and multi‑model detection
  • Benefits of a PC‑free, low‑cost and scalable solution

Video timeline

  • 00:00 - Introduction
  • 00:26 - Company overview
  • 01:42 - Object localization
  • 04:28 - PC implementation
  • 04:50 - New camera
  • 05:35 - Final product
  • 06:53 - Three steps
  • 07:45 - Example
  • 10:58 - Software
  • 16:28 - Programming
  • 19:44 - Questions

Who should watch

Ideal for robot developers, automation engineers, system integrators and manufacturers seeking an accessible, cost‑efficient and AI‑powered pick and place solution that requires minimal setup.

Watch now to see how IDS NXT and EyeBot make vision‑guided pick and place simpler, smarter and more flexible.

Discover products
Further information
Contact
Speaker Alexey Pavlov next to robotic arm and IDS uEye camera
Alexey Pavlov
CEO

urobots GmbH

Similar content

Thumbnail from Vision Channel Session " How-To use AI-based inference with IDS peak"
How-to use AI-based inference with IDS peak

Would you like to integrate AI-based image analysis into an existing application without the need for special intelligent cameras or having to familiarise yourself with AI-based image processing?
11 February 2025
Thumbnail from Vision Channel Session "Reliable AI vision works without environment adjustments!"
Reliable AI vision without labs or complex setups

Reliable image processing usually starts with a perfectly staged object environment to ensure stable and reproducible camera images.
26 November 2024
Thumbnail from Vision Channel Session " Build and deploy ai based machine vision systems"
How-to build and deploy AI-based machine vision systems

The use of machine learning operations in image processing applications makes it possible to solve tasks that were previously difficult or only possible with a great deal of effort.
25 June 2024
Thumbnail from Vision Channel Session " How-To read any text reliably with DENKnet OCR"
How-to read OCR with DENKnet for industrial applications

Would you like to know how reliable machine reading of any type of character string is possible?
19 March 2024
Thumbnail of Vision Channel Session ‘How-to best create an IDS lighthouse dataset’
How-to create a high-quality image dataset for object detection

Let's talk about datasets and what needs to be done to get the best results for AI vision applications!
22 December 2023

Knowledge base

Robot plays “Rock, Paper, Scissors” - Part 3/3

In part 3 of our serialised story, we play: The robotic arm plays rock, paper, scissors with a human.
A vision app as a simulated computer opponent

Part 2 of our sequel deals with the further processing of recognised gestures using the vision app as an AI opponent.
Gesture recognition with intelligent camera

Sebastian Trella is passionate about technology and robotics. In his own blog, he is always taking on new tasks.
Back to the overview