By Trine Reinholt Andersen, AAU Technology Transfer Office

Robotic welding has become an indispensable solution in the manufacturing sector, ensuring precision, quality consistency, and higher throughput. In Denmark, using robotics is the natural path to take in a high-cost environment where skilled welders are scarce and very expensive. However, programming these robots remains a complex and time-consuming task meaning that smaller companies with a need for one-of-a-kind manufacturing, or small-scale production series, often refrain from using robotics. Furthermore, current solutions involve a high degree of skill and subject matter expertise, requiring specialist engineering competences, making them unattainable by small and medium-sized enterprises (SMEs).

The lack of seamless integration between Computer Aided Design (CAD) software and robotic systems further exacerbates the problem. Human operators often must bridge this gap manually by interpreting design files and translating them into executable instructions for the welding robots, taking up much unnecessary time. Additionally, ensuring quality and accuracy in complex welds, particularly for irregular or customized parts, demands significant expertise and effort to get the offline programming done right. These challenges can result in delays, increased costs, and limited scalability for manufacturers seeking to adopt robotic welding solutions.

AT THE FOREFRONT OF WELDING ROBOT PROGRAMMING

RoboSynapse, a spin-out from Aalborg University, specializes in digital integration of welding robot programming with a focus on adopting the tools SMEs already have access to. This eliminates the need for manual transfer of welding information, thus, making robotics for welding a much more attainable solution for SMEs and more attractive in general for companies which have a high volume of robot programming tasks. 

Traditionally, design engineers create the 3D CAD models of the products and then print 2D drawings of the CAD models which contain weldment annotations as per ISO 2553:2019, including geometry, size, and welding process information. These annotations are displayed on the printed technical drafts and processed manually by the robotics engineers or the welders. 

As a result, the robotics engineer must manually reprogram weldment details using the printed technical draft, duplicating the design engineers' work. This process increases lead time, adds redundancy, reduces welding-robot utilization, and increases the risk of error due to manual processing of data.

By leveraging cutting-edge automation, RoboSynapse’s solution can extract all the data required to program a welding robot directly from the 3D models annotated as per ISO2553:2019 without the need of printing the weldment details. The extracted data from the CAD models are seamlessly imported into the welding robots OLP within a few minutes, enabling the robot to start producing products within minutes rather than days.

With weldment information directly extracted from the CAD models, RoboSynapse’s solution leverages the weldment string annotations precisely as created by the design engineers - typical products contain 20-200 weldment strings.

The solution is compatible with any CAD software meeting ISO 2553:2019 standards, saving valuable time for engineers and eliminate redundancy across the value chain while optimizing the usage of skilled labor.

BASED ON AN AWARD-WINNING PHD PROJECT

RoboSynapse’s basic CAD OLP middleware software is based on a PhD project by former PhD student, and founder of RoboSynapse, Ioan-Matei Sarivan from the Department of Materials and Production at Aalborg University, under the supervision of Professors Brian Vejrum Wæhrens and Ole Madsen.

Matei’s Ph.D. project was awarded the first prize for excellence in research that delivers workable results and solves specific challenges in the Danish manufacturing industry, by the Otto Mønsted’s Foundation’s MADE Award in 2022.

The PhD project was supported by MADE – Manufacturing Academy of Denmark, with Sjørring Maskinfabrik as the main partner due to a need to weld complex, best-in-class buckets and attachments for construction equipment. With a clear need, and the collaboration from Sjørring, Matei was able to create the automation software and demonstrate it both in the lab at AAU and at Sjørring. 

Based on the results, it was clear to Sjørring Maskinfabrik that the solution not only filled a competencies gap but also could assist the company in scaling and increasing productivity. Hence, the establishment of RoboSynapse as a company was encouraged by Sjørring who further was ready to purchase the CAD OLP middleware software and become the first customer.

THE COMMERCIALIZATION PROCESS

The AAU Technology Transfer Office has been part of the process from supporting Matei on the entrepreneurial journey, securing the interest, and clearing all IP rights with the involved partners to secondly, working with Matei to uncover the commercial potential of the software, preparing the business model, and strengthening the team formation towards the establishment of RoboSynapse as a spin-out in 2024 and later the licensing of the software from Aalborg University to RoboSynapse ApS.

During the process of forming the team around RoboSynapse, two graduates from Aalborg University joined Matei in his entrepreneurial journey, including Max Hildebrand, M.Sc. in Manufacturing Technology and in charge of the development tasks, and Alina Seulean, M.Sc. in Business Economics, Culture and Communication and in charge of business development.

While the work at Sjørring continues to accommodate the new digital pipeline on their value chain, RoboSynapse ApS has partnered with a range of CAD software suppliers and OLP software suppliers for development and distribution purposes, including Dassault Systèmes and CENIT. 

RoboSynapse has further received interest from potential customers in Germany, France, Canada, and Mexico.