From Harvard Business Review to the World Economic Forum, there’s one phrase buzzing on everyone’s lips: Industry 4.0—the ongoing digital transformation of manufacturing and production processes across practically all sectors.
Though it only officially entered our lexicon in 2011, Industry 4.0 traces its origins to the first Industrial Revolution in the 18th and 19th centuries. Perhaps the most profound revolution in recorded human history, the combination of water and steam power during the first Industrial Revolution mechanised production at an era when most things were handmade. Since then, the respective adoption of electricity and computers during the second and third Industrial Revolutions have created the automated mass production factories of today.
human involvement. To bring such smart factories into the mainstream, researchers at NTU are collaborating with prominent companies to develop cutting-edge solutions for the unique problems they face in industry. Here’s how NTU is taking Industry 4.0 from buzzword to reality.
Troubleshooting before trouble begins
One such corporate laboratory ushering in the future of industry is the Singtel Cognitive and Artificial Intelligence Lab for Enterprises (SCALE@NTU), set up by NTU and Singaporean telecommunications conglomerate Singtel in 2017 on the NTU campus. By leveraging emerging technologies like AI, data analytics, robotics and smart computing, SCALE@NTU seeks to develop applications and smart urban solutions in areas like manufacturing.
Take the problem of maintenance. Should a particular machine breakdown, production becomes delayed, resulting in unpleasant ripple effects like lowered outputs and missed deadlines, compromising customer trust. To prevent this from happening, SCALE@NTU Deputy Director Assoc Prof Yeo Chai Kiat leads the condition-based maintenance (CbM) programme.
detect a system’s faults, predict instances of future failures as well as automatically suggest maintenance plans.
“In CbM, anomalies can be detected by comparing a machine’s digital footprint against another one operating normally. By collecting and analysing this data, we can use AI to help us decide when it is the best time for maintenance before failure occurs,” explains Assoc Prof Yeo. This is where CbM differs from predictive maintenance, where maintenance is scheduled and performed once a certain threshold of time or usage is achieved.
Accordingly, predictive maintenance can sometimes be too late or too early, causing service downtime or unnecessary maintenance respectively. AI—with its capacity to make sense of the overwhelming deluge of real-time machine insights and historical data, and its ability to extract actionable information—is therefore key to CbM.
Currently, the SCALE@ NTU team has deployed CbM to manage energy use in smart buildings through a platform known as ToAI from Singtel’s subsidiary, NCS. A lightweight approach to building AI-powered applications for the Internet of Things, ToAI also holds vast potential in manufacturing, notes Assoc Prof Yeo.
“Various sensory devices installed on machines can thus be easily connected to monitor and predict operation conditions. This could improve the visibility on a machinery’s operational status, detect early signs of failure and increase equipment uptime and utilisation,” she says.
Not only that, by pre-empting breakdowns before they even occur, CbM solutions are considerably more cost-effective compared to other forms of maintenance. Moving forward, Assoc Prof Yeo’s team at SCALE@NTU is also looking to develop image and video analytics for better workplace safety as well as an integrated planning framework for resource allocation in a complex environment.
sectors. Ultimately, it is part of the many cogs realising digital transformation in Industry 4.0,” Assoc Prof Yeo notes.
The rise of robotic assistants
While CbM helps predict and prevent future machine failures, robots could be the key to fending off fatigue in human workers. First introduced in factories in the 1960s, robots have since become a mainstay worldwide for their ability to handle some of the more dangerous or mundane activities previously performed by humans. Unlike their flesh-and-bone counterparts that require regular breaks from work, machines can handle repetitive tasks with ease.
Still, despite the latest developments in automation, most tasks in industrial sites still require a human touch. After all, human workers can adapt remarkably well to unpredictable situations compared to machines. At NTU’s School of Mechanical and Aerospace Engineering, Assoc Prof Dino Accoto seeks to merge the untiring nature of robots with the flexibility of humans—harnessing the best of both worlds.
To this end, he and his collaborators are developing wearable robots that wrap around the bodies of users like the exoskeletons of animals like insects, enhancing their strength and endurance. In 2020, his team developed a voice-activated robotic exosuit that provides increasing levels of assistance, reducing muscular effort up to 60%.
“In industrial settings, exosuits can help tackle ergonomic and safety issues, making tasks more manageable and less demanding from a physical standpoint,” he explains. “This is particularly important when heavy power tools or other specialised equipment must be handled for some time.”
and endurance, the active life span of an industrial worker can be extended, with an overall positive impact on the sustainability of an ageing society,” adds Assoc Prof Accoto.
Another project of Assoc Prof Accoto’s involves the development of “cobots”—mobile collaborative robots—for smart mobility and delivery. Consisting of a robotic arm mounted on top of a wheeled robotic base, the cobot is meant to cooperate with human operators in industrial settings, picking up how operators perform new tasks and becoming increasingly skilled over time.
Similar to how video tutorials and online courses make knowledge accessible to anyone with an internet connection, the skills gained by one cobot will be made available to other cobots across the world. “Their knowledge-sharing capability will exponentially narrow down the learning curve,” notes Assoc Prof Accoto. “This will lead to a fleet of skilled and ready-to-use cobots immediately available for deployment in different industries.”
Ultimately, for Assoc Prof Accoto, the crucial role of robotics will only become more apparent as developed countries worldwide grapple with the consequences of ageing societies.
“Robotics can increase the output per capita, help to extend working life as well as preserve technical knowledge and manual skills now possessed only by pensionable operators,” he concludes. “Robotics is key to improving our quality of life, by enabling advanced services and products for mobility and delivery, which are central in the future vision of smart cities.”
Release the delivery robots!
Teaming up with German automotive manufacturing company Continental at their namesake Continental-NTU Corporate Lab—established on campus just this April 2021—Prof Boh is embarking on a project to evaluate the usability and adoption of the company’s autonomous delivery robot, Corriere, as it navigates traffic, takes lifts and whizzes through building security.
is, the final step in the process, where a package is moved from a transportation hub to a parcel locker, pick-up point or doorstep. By optimising last-mile delivery, items can be delivered faster and with considerably less logistical frustration.
“We are attempting to understand the underlying business models suitable for this technology, with the aim of broadening the understanding, acceptance and adoption of automated delivery robots in the context of grocery and food delivery,” explains Prof Boh.
“Our objective is to examine and compare the reactions and acceptance of three key stakeholders within the setting where the robots will be deployed,” she adds. These stakeholders are the consumers and merchants who respectively buy and sell the products, as well as the passers-by who share common spaces like walkways and lobbies with the robots en route to their deliveries.
Aside from the practicality of deploying the robot, Prof Boh and her team are also factoring in the emotional experience of stakeholders as they interact with the machine—specifically, the motivators that drive individuals to adopt or accept the new technology, as well as the factors that lead users to reject the robot.
“Amid the global pandemic, such a study is timely and significant as it would help inform the design and deployment of last-mile delivery robots bringing essential goods to consumers safely and hygienically with minimal human contact,” Prof Boh shares.
From proactively maintained machines to robots seamlessly synced with humans, various sectors ranging from manufacturing to logistics have much to look forward to in the next few years. And as NTU researchers work to smoothly deploy these technologies in industrial settings, we can all rest easy knowing that the future of robotics is in good hands.