ISMR December 2023 / January 2024 | Page 13

INDUSTRY NEWS

Core themes for MACH 2024

The Manufacturing Technologies Association ( MTA ) in the UK aims to
MACH 2022 in the UK . spearhead growth by using its flagship event to launch its new Knowledge Hubs initiative . The MTA , which organises the MACH trade show on behalf of the engineeringbased UK manufacturing industry , has been campaigning for greater adoption of new technology for some time .
It will expand upon this at MACH 2024 , which will be held at the NEC in Birmingham ( UK ) from 15-19 April 2023 , by explaining of the exhibition ’ s key themes , especially that implementing the latest techniques in Energy Efficiency , where £ 960 million manufacturing processes is the fastest way has been earmarked for clean energy to boost the UK ’ s output . The programme manufacturing through a ‘ Green Industries showcases the latest technology to
Growth Accelerator ’,” outlined the MTA . exhibitors , helping manufacturers learn
The focus for this will be a series of new more about the potential offered by such Knowledge Hubs to educate manufacturers technology as well as when to adopt it and in when and how to adopt new technologies . how to implement it to best effect .
The hubs , which will have dedicated stands
“ The initiative has been handed a major within the various exhibition zones , will each shot-in-the-arm with the announcement by focus on a particular type of technology : UK Chancellor Jeremy Hunt of £ 4.5 billion of Automation and Robotics ; Data and Artificial funding to support advanced manufacturing . Intelligence ; Energy Efficiency ; Additive The funding will directly benefit several
Manufacturing and Tooling .
In recognition of the importance placed on these hubs , each is being managed by one of the specialist centres from the High Value Manufacturing Catapult such as the Manufacturing Technology Centre ( MTC ) and the Advanced Manufacturing Research Centre ( AMRC ). The Catapult network is widely recognised for the research and development work conducted at its various centres .
Attracting 26,000 visitors and more than 600 exhibitors , MACH connects UK manufacturing engineers ; decisionmakers ; buyers and specifiers with suppliers of new technology ; equipment ; services and processes . It showcases sustainable , innovative technologies used across the manufacturing spectrum . The exhibition is currently more than 85 % sold . Co-located at MACH 2024 will be subsidiary exhibition , The Engineering Supply Chain Show . n
www . machexhibition . com
www . mta . org . uk

AI network vulnerability to malicious attacks

Artificial intelligence ( AI ) tools [ are promising ] for applications ranging from autonomous vehicles to the interpretation of medical images , writes Matt Shipman . However , a new study finds that these AI tools are more vulnerable than previously thought to targeted attacks that effectively force AI systems to make bad decisions .
At issue are so-called “ adversarial attacks ”, in which someone manipulates the data being fed into an AI system to confuse it .
The new study from Tianfu Wu , co-author of a paper on the new work and an associate professor of electrical and computing engineering at North Carolina State University ( USA ), and his collaborators focused on determining how common these sorts of adversarial vulnerabilities are in AI deep neural networks . They found that the vulnerabilities are much more common than previously thought .
“ What ’ s more , we found that attackers can take advantage of these vulnerabilities to force the AI to interpret the data to be whatever they want ,” Wu added . “ This is incredibly important because if an AI system is not robust against these sorts of attacks , you don ’ t want to put the system into practical use ; particularly for applications that can affect human lives .”
To test the vulnerability of deep neural networks to these adversarial attacks , the
researchers developed a piece of software called QuadAttacK . The software can be used to test any deep neural network for adversarial vulnerabilities .
“ Basically , if you have a trained AI system and you test it with clean data , the AI system will behave as predicted . QuadAttacK watches these operations and learns how the AI is making decisions related to the data . This allows QuadAttacK to determine how the data could be manipulated to fool the AI . QuadAttacK then begins sending manipulated data to the AI system to see how the AI responds . If QuadAttacK has identified a vulnerability , it can quickly make the AI see whatever QuadAttacK wants it to see .”
In proof-of-concept testing , the researchers used QuadAttacK to test four deep neural networks : two convolutional neural networks ( ResNet-50 and DenseNet-121 ) and two vision
Image : Shutterstock . com . transformers ( ViT-B and DEiT-S ).
“ We were surprised to find that all four of these networks were very vulnerable to adversarial attacks ,” Wu highlighted . “ We were particularly surprised at the extent to which we could fine-tune the attacks to make the networks see what we wanted them to see . “ Now that we can better identify these vulnerabilities , the next step is to find ways to minimize those vulnerabilities . We already have some potential solutions but the results of that work are still forthcoming ,” Wu concluded . The paper , “ QuadAttacK : A Quadratic Programming Approach to Learning Ordered Top-K Adversarial Attacks ,” was presented on 16 December 2023 at the 37th Conference on Neural Information Processing Systems ( NeurIPS 2023 ) held in New Orleans ( USA ). First author of the paper is Thomas Paniagua , a Ph . D . student at NC State . The paper was co-authored by Ryan Grainger , a Ph . D . student at NC State .
The work was done with support from the U . S . Army Research Office and the National Science Foundation in the U . S .
The research team has made QuadAttacK publicly available , so that the research community can use it themselves to test neural networks for vulnerabilities . n
www
. ncsu . edu
ISMR December 2023 / January 2024 | ismr . net | 13