Fourth industrial revolution under the spotlight at Armscor Engineering Day

499

The fourth industrial revolution and its impact on the aerospace and defence industries came under the spotlight at Armscor’s recent Engineering Day.

The 2019 edition of the Engineering Day was held on 14 August at the Council for Scientific and Industrial Research (CSIR) and was well attended by the industry, with nearly a dozen presenters speaking on topics ranging from the future of work to designing better ammunition.

Dr Hildegarde Koen, a Senior Researcher at the CSIR, gave an overview of the fourth industrial revolution as something that is blurring the lines of the physical, digital and biological spheres. It is being driven by artificial intelligence (AI), robotics, the Internet of Things (IoT) and big data. She explained that big data is when there is so much data available that only computers can effectively make sense of it, gain insights and use it to make better decisions.

The IoT connects multiple sensors and systems into a single environment – for example smart cities can track people through their cellphone signals, numberplate recognition and facial recognition, for example, making transport simpler and security better.

Koen said that in the realm of artificial intelligence, the fourth industrial revolution is being driven by machine learning and deep learning. Machine learning occurs when computers are given labelled data – for instance they can learn to distinguish between fruit by analysing thousands of labelled pictures of fruit. Applications including facial recognition and voice-activated assistants such as Siri. Deep learning occurs when artificial intelligence uses neural networks to process information the same way the human brain does. It is more complex than machine learning and requires more computer power.

In the aerospace and defence environment, artificial intelligence is being used in autonomous vehicles, to predict failures in manufacturing, to recognise and classify targets, maintenance diagnostics, and cyber security (for example wargames/simulation and analysing suspicious digital activity).

Professor Daniel Mashao, Executive Dean, Faculty of Engineering and Built Environment at the University of Johannesburg, added to this by saying the fourth industrial revolution is also driving robotics, unmanned vehicles, exoskeletons, digital/cyber security tools and new materials (especially nano-materials) and manufacturing processes such as 3D printing.

Lisa Abom, Vice President and Chief Technology Officer at Saab Aeronautics, looked at how the fourth industrial revolution has affected engineering in defence by using the Gripen fighter as an example. She said this aircraft was developed digitally from the start, with no paper drawings, and the aircraft was simulated on computers before flying started – flight testing was only to validate simulation data. By digitally designing the Gripen, it is much easier to make changes without requalifying the aircraft, and update it as well – Saab regularly releases software upgrades for the Gripen. Looking to the future, she said the Gripen will one day be flying in a team with unmanned aircraft.

Future trends Abom identified include more sensors, increased computational power and high-energy weapons on aircraft as well as increased manned/unmanned aircraft interaction.

Looking at South African developments, Kyle Winnaar from Flamengro said the company had developed cheap and robust test and evaluation software using digital image processing and machine learning for displacement and acceleration measurement. This is used in measuring the properties of armour, materials and weapons. It has been tested in analysing weapon feed chute dynamics, the fragmentation of a prototype warhead and projectile tracking to determine muzzle velocities.

Flamengro’s Adriaan Steenkamp outlined the use of fourth industrial revolution technologies (including 3D printing, a digital twin and smart sensors) to assist with the development of a 30 mm projectile for air defence. This was developed and qualified in 18 months for less than R100 million. The idea was to take an off-the-shelf round and improve its air defence capabilities, with the projectile designed to destroy incoming mortars, rockets, artillery, helicopters and unmanned aerial vehicles, with fewer rounds needed to destroy a target than existing gun-based air-defence systems. Tests were successfully conducted against 120 mm mortar and 155 mm artillery rounds.

Flamengro detailed another of its projects, with Pitso Magoro explaining how machine learning on a digital twin (digital replica) of a turreted cannon was used to diagnose failures and improve reliability. Magoro said that automatic cannons often experience jamming and bunching of ammunition, with the fast movement of the ammunition belt causing problems such as cartridge cases catching on the feed mechanism and rounds bunching up during feeding.

Through the digital model, a retaining device was added to mitigate bunching whilst the cannon’s internal geometry was improved to reduce jamming. Magoro said that these problems were solved cheaply by creating the digital models.

Sicelo Ngubane, Managing Director of MND Technologies looked at a technology that has been enabled by advances in computer power: passive radar. This uses electromagnetic signals emitted by illuminators of opportunity (such as FM radio towers, digital TV signals, cellphone towers etc.) to detect targets. Conventional radars have their own transmitter and receiver, but passive radars only have a receiver as they rely on other illuminators. This makes them cheaper, less complex (no rotating parts) and able to detect stealthy aircraft. It is also very difficult to jam or spoof passive radars and nearly impossible to detect them.

Barry Vorster, a partner at PricewaterhouseCoopers South Africa, gave a presentation on the fourth industrial revolution and the future of work. He explained that new technologies are changing employment and making people work harder as technology allows them to become more productive. Approximately 65% of children entering primary school will hold jobs that currently don’t exist today, with the future of work being influenced by artificial intelligence, augmented reality, blockchain, drones, the Internet of Things, robotics, virtual reality and 3D printing.

Looking to the future, Vorster said that as people live longer, they can be expected to work into their 70s and 80s and be constantly learning new skills to keep pace with developments. Jobs will disappear in the manufacturing, wholesale, retail and finance industries as automation takes over, and humans will become more valued for things like creativity, persuasion and innovation.



Also touching on the future of work, Cobus Oosthuizen, CEO Siemens Industry Software South Africa, predicted that robots will displace 75 million jobs globally by 2020 but create 133 million new ones – a net positive. He said that technology will not be a replacement for many of the skills needed for the jobs of the future but it will act as a supporter and integrator making digital fluency just as important as literacy.