The basics of vision guided robotics

The evolution of Vision Guided Robotics

Robots are very common nowadays in our lives. The time where everything had to be done manually, are rapidly shifting towards a world where everything can be programmed or managed automatically. Prime examples of this are washing machines, vacuum robots or a drone that is capable of flying an recording autonomously.

Of course, it wouldn’t take long for these developments to find their way into industrial applications. A well known example of robotics within the industry is the use of robot arms for the assembly of cars or for welding constructions.

Robotarm car industryRobotarm in car industry

But these robots came with a massive constraint; they couldn’t see. The robots were totally dependent of software written for each individual task. Every new or variation of a task, such as moving a product would require a total rewrite of the script or software.

This all changed with the rise of vision systems for robots. A robot connected to a camera system or a smart viewer could now detect objects and perform different tasks based on those objects. The simple robot arms mentioned earlier could now be transformed into vision guided ‘smart robots’. Despite the fact that smart vision technology is around for many years, as shown by Teqram’s solutions, this technology is still rapidly evolving with high expectations for the future; industry 4.0.

Object recognition: From laser to camera

Our current robots are becomming smarter every day. Normally a robot would be fully dependent on software of scripts for individual tasks, nowadays – a robot is more and more capable of controlling itself through machine vision. With ever expanding vision techniques. A common example of a early application of vision techniques is the use of laserscanners with a barcode. By scanning the barcode, the robot could recognise the product and what value or task was needed for it. This technique is widely used in our everyday life, such as in supermarkets or at industrial plants.

A stap head in this technique is the QR code. The QR code became massively known with the evolution of smartphones. A QR code is most known as a hyperlink, build up from a square with many smaller squares inside it. A camera is capable of reading the QR code, recognising the link and automatically performs a task – such as opening it.

The next step in the evolution of product recognition is autonomous detection of objects with a camera. This technique is being developed since a few years and the possibilities that come with it are ever expanding. A practical example of this is a camera that is able to recognise faces, or a gate at a parking lot that recognises a license plate. Perhaps the biggest opportunities for this technique are within the industry. Due to clever cameras that are able to detect and recognise objects and products, the robot is now capable of performing individual tasks for every product autonomously. Object recognition allows a robot to be used faster and in more effective ways at production sites.

3D vision

The application of 3D camera images are rapidly increasing compared to 2D camera images. By using multiple sensors, such as sensors for surroundings and position, combined with a 3D time-of-flight laser scanner, it is possible to create 3D models of products on pallets in high detail. These 3D scans are created with a time-of-flight laser that measure the time it takes for light to travel to the product. The biggest advantage for this method is that it can be used for every surface. This enables a robot to recognise randomly placed objects within a box or on a pallet and move it onto ordered stacks of the same products, also called Bin-Picking.

Software and algorithms

An essensial part of these advanced vision guided robotics is the software that is able to control the algorithms. These allow the robot to recognise dimensions and prevent it from searching outside these dimensions. The software also calculates which product can be picked from a stack and the amound of force that is required to pick it. Software and algorithms make the robot more safe, faster and more effective.

robot loading conveyor beltRobots loading a conveyor belt

Application of vision guided robotics

Vision guided robotics will play an important role within the industrial automation business. There have been several succesful applications already including:

  • Loading and unloading of pallets onto machines
  • Machine tending; loading and unloading of machines
  • Pick and place robots
  • Approval/disapproval of products and parts
  • Object identification and individual product handling

Apart from these applications there are many other possibilities to implement vision guided robotics for proces automation. By using flexible software and Teqram’s easEye® vision system you can also experience the endless possibilities of vision guided robotics. We kindly help you automate your industrial processes, such as deburring machines.