![IBM-Samsung article](/-/media/imda/images/content/news-and-events/impact-news-2024/08/samsung-use-case/ibm-samsung-article.webp?h=516&w=916&la=en&hash=57D73E2649851CF95D3C504E438B0727)
IMDA 5G Use-Case Findings
Industry 4.01 represents a change beyond just an interconnected and autonomous digital enterprise. Here smart connected technologies are embedded with artificial intelligence (AI)2, cognitive technologies3 and integrated with digital twins4, to name a few examples. With Industry 4.0, enterprises are able to communicate, analyse and use data to produce more efficient and productive processes for smarter manufacturing.
In Singapore, our vision for Manufacturing 2030, is to become a global business innovation and talent hub for advanced manufacturing. Today, manufacturing accounts for 20% of GDP and Singapore is looking at growing this sector by 50% of its current value of S$106billion5.
This 5G Industry 4.0 trial was a global first for both IBM and Samsung. Such a collaboration allowed for the development, testing and innovation of 5G-enabled solutions with AI and machine learning, across the manufacturing sector. It also served as an opportunity for 5G standalone trial and for all parties to explore how hardware and network could be optimised to meet the needs of the advanced manufacturing sector.
The objective of this trial was to test various use cases, encompassing visual analytics, acoustic analytics, and augmented reality (AR) across the manufacturing sector, with the availability of Enhanced Mobile Broadband (eMBB6) and low latency7 5G networks. With Singapore accelerating 5G standalone coverage, it is important to evaluate possible use cases, and optimise equipment and software to harness the true potential of 5G in order to support the nation’s move towards a smart nation.
A comparison between 5G and Wi-Fi 5 was also performed to understand the reduction in investment cost, performance and communication reliability parameters. Both organisations planned to utilise technical learnings and benefits attained from this trial in their global deployments.
Optimising productivity, efficiency, and reducing operational costs, were key drivers for this digital transformation within Industry 4.0.
Joan Yeo
Head of Department[br/]5G Industry 4.0 Development, IBM
Automation with intelligent visual and acoustic analytics
For IBM, one of the key areas of enhancement within their Industry 4.0 portfolio, was to compare the efficiency of hosting a centralised server for video analytics8 and AI processing. The trial demonstrated that by using mobile devices connected over 5G, it could handle the necessary bandwidth for video data transmission back to a centralised server. “Scaling out Industry 4.0 solutions can be very costly and expensive, with 5G we were able to centralise our compute resources to reduce the cost of intelligent edge devices” explains Joan. “When looking at the infrastructure architecture level, it also changes the way solutions are deployed,” says Christian Nugraha, Cognitive Solutions Architect at IBM.
In addition to data bandwidth, assured network latency9 was another critical parameter for consideration, as real-time interaction was important. IBM tested this concept, by congesting the network to prove that, 5G could also deliver on this promise. The video analytic solution was able to continue to detect and track objects of interest, during network congestion with 5G. However, with Wi-Fi, the congested network caused the video to freeze and overall solution to eventually fail. Wi-Fi on a congested network, had significant impact on the frame ingestion towards the deep-learning engine, causing mis-tracking and inability to process video feeds.
During the COVID-19 pandemic, the manufacturing sector had a need to put in place social distancing. This coupled with requirements to wear Personal Protection Equipment10 (PPE) to ensure overall worker safety, was paramount for continued business operations. IBM recognised the opportunity to extend its solution to address both these use cases. Realtime visual recognition and video analytics were used to ensure workers wear their PPEs before entering the work floor. Whilst object detection and tracking, were used to ensure workers comply to safe distancing rules. These use cases, were further expanded to support the healthcare industry. Robotic surgery, intelligent video analytics and remote telemedicine, are examples where high bandwidth video streaming and low latency 5G networks, become key enablers for new processes. This is also supported by Samsung’s advanced camera found on their Galaxy smartphones, where high quality captures add to efficiency of IBM’s visual recognition and analysis solution, realising real-time applications on factory floors.
In a separate use case with acoustic analytics11, binary streaming12 was used to transmit audio to an acoustic engine application server, for intelligent sound classification. Using this acoustic AI application, automation of product inspection helped to increase manufacturing yield and quality assurance. Sound analysis was used to identify defects in finished products, monitor and track anomalies, which may not be visible to the human eye. It was also used in preventive maintenance, to determine when a machine or part will need to be replaced or repaired. Here 5G played an important role, in maintaining consistent audio ingestion to the acoustic inference engine13 and ensuring real-time alerts during network congestion.
When Wi-Fi was used over a congested network, there were inconsistent audio streams which affected sound classification. This resulted in end-to-end real-time alerts of 18s on Wi-Fi during network congestion, as compared to 35ms on 5G. A further breakdown of notification latency from server to end user devices, showed a degradation from 11.6ms (average non-congested Wi-Fi) to 5,135.4ms (average congested Wi-Fi), which meant timely notifications were not guaranteed. However, with 5G the latency during congestion was insignificant with a slight degradation from 11.8ms (average non-congested 5G) to 14.1ms (average congested 5G).
With a mobile visual and acoustic analytics capability, IBM is looking to expand its use across the building and construction sector, for improved efficiencies and automating site inspection.
Invaluable insights through augmented reality
Augmented Reality (AR)14 models have previously been preloaded onto user devices, making this process cumbersome, as these files are large and change over time. With new features like dynamic AR models, these no longer need to be preloaded, allowing for mainstream mobile tablets with high-speed connectivity to be adopted. Here the layering of digital information, can now be rendered at the backend and streamed directly to user devices over 5G. With WebRTC, video communications, shared 3D object visualisation and live annotations create a seamless collaborative platform between an expert and field engineer. IBM tested such AR features with mobile tablets, for operational insights and to guide engineers on the production floor.
With 5G, a 1.2GB AR model could be directly streamed in half the time as compared to Wi-Fi. Furthermore, during network congestion, Wi-Fi experienced an increase in video delay from 160ms to 448ms, while 5G showed minimal impact. Overall network congestion on Wi-Fi, hindered the imperative use of dynamic AR models and left behind a degraded video user experience.
“With AR deployed across Industry 4.0, we are looking at improving new hire training and efficiency by 50%, reducing training costs by 20%, and improving productivity by 10%,” says Joan. The future of AR has relevant applications across security solutions, specifically in the field of forensic analysis where the overlay of digital information onto the real world is vital.
Empowering Industry 4.0 solutions with 5G-ready connectivity and devices
Advanced smart manufacturing solutions need to be backed with suitable connectivity and devices. This is where Samsung comes in. As a leader in delivering end-to-end 5G solutions, Samsung worked closely with IBM in delivering 5G hardware solutions such as Galaxy smartphones and antenna modules that help reduce latency, making smart manufacturing scenarios a reality.
“5G allows manufacturing organisations to unlock the true potential of advanced technology to create new use case scenarios that can have a real impact in realising the vision of smart factories. With the sector exploring and deploying more new technology, there will be a need for a reliable 5G infrastructure and hardware partner who can offer the backbone for industrial automation in a wireless scenario,” says Timothy Tan, Head of Enterprise Business, Mobile Experience, Samsung Electronics Singapore.
The trial has demonstrated positive initial results, which Timothy shares that “it’s a validation of our 5G capabilities.” Samsung continues to work with industry partners and organisations on trials in the advanced manufacturing sector, with a goal of supporting its customers and partners in Singapore in realising real business benefits of digitisation.
Realising the vision of Industry 4.0
This trial also led to the birth of Singapore’s first 5G Industry 4.0 Studio15. The Industry 4.0 Studio, is meant to innovate, test, deploy 5G-enabled use cases for advanced manufacturing with a spotlight on: automated visual inspection using AI for image recognition and video analytics; improved equipment monitoring and predictive maintenance using AI-enabled acoustics insights and using AR to improve productivity and quality assurance. “With 5G, came the opportunity to improve our architecture stack, change our deployment model and innovate on Industry 4.0 use cases”, shares Joan.
Following a successful trial, the 5G Industry 4.0 Studio has since been replicated across other IBM Global sites such as IBM’s Dallas Client Center, and IT manufacturing use cases deployed at IBM’s own Poughkeepsie, New York manufacturing plant.
Moreover, the combined efforts between IBM and Samsung will facilitate enterprises to swiftly customise and deploy their applications, while flexibly managing them across any cloud, on-premise or private environment of their choosing.
Footnote
1 Industry 4.0 or the Fourth Industrial Revolution (4IR): has been defined as technological developments in cyber-physical systems such as high capacity connectivity; new human-machine interaction modes such as touch interfaces and virtual reality systems; and improvements in transferring digital instructions to the physical world including robotics and 3D printing; the Internet of Things (IoT); “big data” and cloud computing; artificial intelligence-based systems; improvements to and uptake of Off-Grid / Stand-Alone Renewable Energy Systems.
2 Artificial intelligence (AI): is the ability of a computer or a device controlled by a computer to do tasks that are usually done by humans because they require human intelligence and discernment.
3 Cognitive technologies: is a field of computer science that mimics functions of the human brain through various means, including natural language processing, data mining and pattern recognition.
4 Digital twin: is a virtual representation of a real-world physical system or product that serves as the indistinguishable digital counterpart of it for practical purposes, such as system simulation, integration, testing, monitoring, and maintenance.
5 See EDB reference article (Feb 2021).
6 Enhanced Mobile Broadband (eMBB): is one of three primary 5G New Radio (NR) use cases defined by the 3GPP. eMBB are data-driven use cases requiring high data rates across a wide coverage area.
7 Low latency: describes a computer network that is optimised to process a very high volume of data messages with minimal delay (latency). These networks are designed to support operations that require near real-time access to rapidly changing data.
8 Video analytics: a type of technology that automatically analyses video content. This is done using algorithms that process video to carry out a specific task - for example, identifying moving objects or reading vehicle licence plates. When artificial intelligence is involved, they are often referred to as intelligent video analytics.
9 Network latency, or lag, is the term used to describe delays in communication over a network. In networking, it is best thought of as the amount of time taken for a packet of data to travel through multiple devices, then be received at its destination and decoded.
10 Personal protective equipment (PPE): is equipment worn to minimize exposure to hazards that cause serious workplace injuries and illnesses. These injuries and illnesses may result from contact with chemical, radiological, physical, electrical, mechanical, or other workplace hazards. Personal protective equipment may include items such as gloves, safety glasses and shoes, earplugs, hard hats, respirators, or coveralls, vests, and full body suits.
11 Acoustic analytics: a type of technology that automatically analyses audio (sound or noise levels). This is done using algorithms that process audio to carry out a specific task - for example, identifying defective objects or detecting anomalies. When artificial intelligence is involved, they are often referred to as intelligent acoustic analytics.
12 Binary streaming: contain a sequence of bytes. In binary streams, there are no translations of any characters on input or output. It treats them as a continuous stream of bytes and ignores any record boundaries.
13 Inference engine: is a component of the system that applies logical rules to the knowledge base to deduce new information. The first inference engines were components of expert systems. The typical expert system consisted of a knowledge base and an inference engine.
14 Augmented Reality (AR) is the layering of digital information on top of the real world where we live in. In Industry 4.0, AR can provide the information engineers need, right on the production floor.
15 Industry 4.0 Studio: IBM’s Industry 4.0 facility in Singapore.