RESEARCH NEWS
Manufacturing / engineering investment
Leicester College in the UK will formally open its Advanced Manufacturing and Engineering facilities at the Abbey Park campus following an exciting multi-millionpound investment in resources , equipment and teaching spaces . This will enable the college to provide industry-standard technical training to students and employers across the region .
Leicester College is one of the largest providers of advanced manufacturing and engineering education in Leicester and Leicestershire . The project was supported by capital funding via the T-Level capital fund .
“ The upgraded facilities will improve the practical and technical experience for existing and future students , studying for T-Level qualifications in Design and Development for Engineering and Manufacturing ( Electrical , Electronic and Mechanical Engineering ) and Maintenance , Installation and Repair for Engineering and Manufacturing , together with a range of other programmes including new Higher Technical Qualifications ( HTQs ) in Engineering . They will also enable the college to continue to support regional employers to develop the skills and knowledge needed to develop the engineering workforce of the future ,” commented Leicester College .
The project included the installation of
Images : Beth Walsh .
engineering laboratories and workshops to develop skills in robotics ; programmable logic controllers ; electronic circuits ; electrical systems ; mechanical systems ; hydraulic and pneumatic systems together with a range of mills ; lathes ; grinders ; rigs ; testing equipment and simulators . The new facilities were designed by Moss Architecture Interiors Ltd and constructed by regional contractor , Stepnell .
Ibrar Raja , Director of Engineering , Leicester College said : “ With the introduction of T Levels and the transition to HTQs , the Higher Education offering for engineering students is undergoing a shift towards addressing future recruitment issues , and local and national skills gaps . The funding received has enabled us to invest in our infrastructure for students , ensuring that we can provide them with real pathways to progression within their chosen engineering sector and safeguarding the skillsets for future generations .”
Leicester College offers engineering programmes from foundation level 1 , for those initially developing their engineering journey , through to higher level qualification including T- Levels , Access courses and foundation degrees . Students can specialise in electrical ; electronics ; manufacturing ; aeronautical ; mechanical ; electromechanical ; space engineering ; technical support engineering ; fabrication and welding ; electric , hybrid and hydrogen vehicle technologies .
Dedicated employer training can also aid upskilling of the workforce to meet local , regional and national industry needs . n
www
. leicestercollege . ac . uk
AI utilisation for welding error detection
Saving raw materials and energy in production processes is key . The same goes for welding . Artificial intelligence ( AI ) can help with this task but the relevant data is needed to train AI systems . However , this is data that many customers don ’ t wish to give away . Federated learning can help to solve this dilemma and Fraunhofer IPA has developed a corresponding AI concept for welding specialist , Lorch .
“ We train the artificial intelligence with the customers ’ data without the data leaving the respective company ,” explained Can Kaymakci , a scientist at Fraunhofer IPA . Each customer trains their own AI model with their data : it is not the data that is exchanged , only the AI models . These are combined into a single , better optimised overall model .
Researchers at Fraunhofer IPA had to select a suitable AI model for energy anomaly detection ( a model that detects user errors primarily through energy consumption data ). To do this , they collected data in the Lorch laboratory about the welding process being observed , including the intentional inclusion of “ user errors ”. They carried out around 200
welding tests . A lot , but not enough to train an artificial intelligence system .
“ We therefore duplicated the data ; the original 200 data sets became 2,200 ,” explained Kaymakci . The team also investigated how many measurements per second are necessary to reliably detect user errors .
“ In this way , we can reduce the required storage capacity , simplify communication and process less data which , in turn , saves time , costs and energy ,” summarised Kaymakci . The researchers implemented the model they created on a welding power source from Lorch . The recognition rate of a model that was trained using federated learning was 0.81 and is therefore comparable to that of a system for which all customer data was available for training . Here , the recognition rate is 0.86 .
“ In contrast , systems that were trained with only one customer ’ s data only detect errors at a rate of 0.45 ,” confirmed Kaymakci . For welding machine manufacturer Lorch , this means that it will be able to offer its customers added value via the AI system without having to store the data centrally at
Lorch . For customers , in turn , there is the advantage of being able to identify errors more quickly and benefit from the “ knowledge ” of all customers . n
www . ipa . fraunhofer . de
In Lorch ’ s laboratory , data from the welding process is collected to train artificial intelligence systems . Image © Lorch .
16 | ismr . net | ISMR May 2024