In the UK, an estimated 744,000m3 of nuclear waste must be safely contained between now and 2140. This monumental decommission is demanding rapid production of containers, however the shortage of skilled welders is just one of the many challenges facing the industry.
With safety considerations being fundamental to all aspects of the nuclear industry, adoption of innovative manufacturing technologies that could drive efficiencies within the sector are slow to be adopted.
One way to de-risk implementation of technologies which could solve these challenges is through government-backed R&D projects.
UK-based metrology company Insphere has developed IONA, a camera system for high-accuracy tracking and guidance of industrial robots. In 2022 they were awarded a Smart Grant by I-UK to address the challenges of robotic welding, in a project named AFFIRM.
Robotic welding has the potential to address shortage of skilled welders, and with advanced robot guidance, automation of container production could reduce production costs, enabling UK supply chains to meet the huge demand, and simultaneously deliver an improved product.
Project AFFIRM
Insphere is the lead partner on project AFFIRM (advanced photogrammetry for flexible intelligent robotic manufacturing), collaborating with Createc and the University of Sheffield’s Nuclear Advanced Manufacturing Research Centre (NAMRC).
The main project outcome is an integrated demonstrator cell, hosted at NAMRC.
To date, Insphere, Createc and NAMRC have integrated key system components into an advanced robot cell in the NAMRC. The cell comprises a welding robot suspended from a multi-axis gantry. A rotary platform is mounted below the welding robot, and a second robot holding a Photoneo sensor is located next to the rotary table.
Insphere’s IONA robot tracking system in the cell comprises six sensors simultaneously tracking the robots and the rotary table. IONA is integrated with the robot controllers, so that its measurement data can be used directly for control and correction of robot position and orientation.
In use, plates to be welded are clamped onto the welding platform; the Photoneo sensor is traversed along the weld line to measure the weld gap; then this dataset is used to plan a robot path for the weld. The welding robot then performs the weld while IONA tracks the true path for validation of weld accuracy and potential correction of the path; finally the Photoneo sensor tracks along the weld line a second time to assess weld metrics and detect any deformation of the steel plates.
The workflow has been successfully tested in the demonstrator cell, and experiments are underway to assess variables including robot accuracy, the influence of environmental variables, and material performance when parameters are varied.
A harsh environment
High temperatures, high-intensity radiation, smoke and sparks all have the potential to hinder tracking and even damage system components. These influences are being rigorously tested, and the NAMRC cell is proving very valuable to maximise the learning from the project.
To date, IONA’s lens filters and software algorithms mask infrared light successfully and weld tracking has been robust and reliable. Long-term reliability tests will continue for the next 12-months.
Machine learning and AI
Robot control can benefit from AI algorithms and IONA provides a rich data source for machine learning. The project is developing deep insights into the effects of influencing parameters, not only to improve individual robot paths, but to progressively drive down errors for all robot programs.
The project will generate machine learning algorithms to deliver a data-driven manufacturing solution, with an anticipated step-change in performance.
Conclusions and next steps
After successful demonstrations of weld tracking and path improvement, the next stage towards commercialisation will be a pilot scheme installing demonstrator cells in customer facilities working within the nuclear sector and beyond.
Companies with challenging or safety critical welding applications are invited to contact the research team to discuss how they can shape project outcomes and participate in the 2024 pilot scheme.
Whilst the direct beneficiaries of successful project outcomes are in the nuclear industry supply chain, wider benefits in the field of automated welding are expected, including in manufacturing for offshore energy, construction of modular reactors, and even in the automotive and aerospace sectors.
For more information about IONA, visit the Insphere website here.
About the author
Craig Davey is Insphere’s Chief Operating Officer. Originally working as a medical engineer, he specialised in quantifying uncertainties of flowrate measurement, and more recently has focussed on large volume metrology applications in manufacturing industries.