Featured Product
This Week in Quality Digest Live
Innovation Features
Ian Wright
MIT and ETH Zurich engineers use computer vision to help adjust material deposition rates in real time
Having more pixels could advance everything from biomedical imaging to astronomical observations
Chris Caldwell
Significant breakthroughs are required, but fully automated facilities are in the future
Leah Chan Grinvald
Independent repair shops are fighting for access to vehicles’ increasingly sophisticated data
Adam Zewe
How do these systems differ from other AI?

More Features

Innovation News
Exploring how a high-altitude electromagnetic pulse works
High-capacity solution using TSMC’s 3DFabric technologies
EcoBell paints plastic parts with minimal material consumption
Study of intelligent noise reduction in pediatric study
Easy to use, automated measurement collection
A tool to help detect sinister email
Funding will scale Aigen’s robotic fleet, launching on farms in spring 2024
High-end microscope camera for life science and industrial applications

More News

David Suttle


Autonomous Robotic Welding

What is breakthrough technology really capable of?

Published: Tuesday, September 26, 2023 - 11:02

You often hear about self-driving cars and their levels of autonomy. When can drivers completely remove their hands from the steering wheel? This also applies to robots. How can robots become fully autonomous?

What are autonomous robots?

Let’s look at the levels of freedom for self-driving cars and draw parallels with the levels of automation in robotic welding.

Level 0: No automation
Just like with self-driving cars at Level 0, everything needs to be done manually. You have a part that must be perfectly pre-assembled without any deviations. Otherwise, robots won’t be able to complete the task. You need to place this part in a precisely defined position, and deviations here are also unacceptable. A highly skilled programmer uses a teach pendant to program every trajectory of the robot’s path.

Level 1: Assistance
At Level 1 of self-driving cars, features like cruise control emerge. Similarly, for robots, individual functions also arise that enable them to adapt to changes in the physical world; for instance, there’s arc sense.

Level 2: Partial automation
At Level 2 a human can take their hands off the steering wheel, but they need to monitor the road. In the case of robots, offline programming comes into play. Primarily, it allows programming to occur not beside the robot but within a virtual environment, reducing robot downtime. It also becomes more convenient for the programmer. However, you still require a highly skilled programmer who will program each trajectory of the robot’s path. Additionally, any change—whether in pre-assembling a part or its placement in the robotic cell—demands programming.

Level 3: Conditional automation
Level 3 represents a significant leap forward. Self-driving cars are already capable of performing most driving tasks, yet a human presence remains necessary for control.

This newfound freedom has been brought to the realm of robotics by ABAGY. The system now completely automates the robot’s path generation; no programming is required at all. Furthermore, ABAGY brings adaptability: A part can already have deviations, and it can be placed freely within the workspace. The company employs machine vision to enable robots to “see,” alongside mathematical algorithms that calculate everything automatically.

Level 4: High automation
Self-driving cars are currently at Level 4. Their level of freedom has already reached 99%. When it comes to autonomous robots, work with them becomes incredibly straightforward. Metaphorically speaking, it’s like speaking in the welding language. You don’t need to be a robotics expert or a programmer; having welding expertise is sufficient. All you need to do is define the welding parameters you require.

Level 5: Full automation
In the case of self-driving cars, reaching the level where one can confidently say, “Zero human attention or interaction,” has proven to be challenging. When it comes to robots, there’s no certainty that manufacturing will become 100% human-free. Frankly, it’s still an open question that makes sense to discuss later, after passing Level 4.

What is breakthrough technology really capable of?

Let’s explore specific examples from real production sites. This will provide the best insight into how breakthrough technologies can truly affect us right now.

Bridge fabrication: Ease of switching from one product to another

The main advantage of adaptive and autonomous robots is the seamless transition from one product to another. This capability allows for the use of robots in high-mix production environments. Here’s the first example.

Conventionally, most bridge structures worldwide are welded manually. Implementing robots in this process presents a significant challenge because each bridge part appears custom from a robot’s perspective. However, with ABAGY, this one cell can weld approximately 50 different products each month. The company can switch from one product to another easily—no programming.

No ‘zero points’

Here, a part is loaded into the working area using a crane. With the traditional approach, you would have to place the part very accurately. And when it’s such a heavy beam, it’s extremely difficult to do this. In the case of autonomous robots, the part is placed freely without any zero points; however, it’s convenient to position. After all, thanks to machine vision, the robots will be able to “see” exactly where it is.

Deformations during welding

Bridge structures are usually huge and can get deformed during welding. This is a telling photo. See how curved the bridge beam is, about 4 in. off. But this isn’t a problem for ABAGY.

Steel conveyor chain: Simple fixtures

Typically, this type of welding is done by hand due to the difficulty of fixing this product in the working area, since the chain consists of many links. In this case, the chain is freely placed on the welding table. However, with the help of machine vision and mathematical algorithms, the robot completes this task with ease.

Excavator bucket

Another example is the excavator bucket. There are numerous variations of these buckets, often customized to specific orders.

Deviations due to metal bending
Even when dealing with a typical digging bucket, every individual bucket essentially takes on a unique identity from the perspective of robots. This is primarily due to deviations caused by the process of metal bending during the pre-assembling.

Part rotation

Additionally, due to its complex shape, the bucket must be continuously rotated. This is a requirement for both a human welder and a robot. In the case of ABAGY, the synchronization of the robot’s movements and the positioner also occurs automatically.


And here’s the rotor. Again, we’re addressing high-mix production. The same complexity applies to rotors—rotor blades might have slight deviations from the perfectly aligned position during assembly, and the robot must be capable of adapting to these changes. Autonomous robots excel at handling these variations, which is crucial in our case.

ABAGY is an AI- and machine vision-based software for industrial robots. This article was published Aug. 15, 2023, on the FABTECH blog.


About The Author

David Suttle’s picture

David Suttle

David Suttle, chief revenue officer at ABAGY, has more than 30 years’ experience in manufacturing tech in leadership positions across a variety of industries and corporations, including Norfolk Naval Shipyard, Reliance Electric, CompuServe Networks, Edison Welding Institute, Path Robotics, and ABAGY.