The age of the fourth industrial revolution is upon us, and world industries are changing fast in accordance. The manufacturing sector, in particular, is seeing a transition to ‘Smart Factory,’ in which technologies like AI, IoT(Internet-of-Things), and ICT(Information Communication and Technology) are applied to the whole product cycle, ranging from product planning and manufacture to sales and marketing.
The reason for the transition is clear: monitoring and analyzing the whole process with digital data and ICT technology allows for a real-time, efficient, quality-assuring system of production that is also flexible enough to react to market demand. Such a system, then, will naturally lead to increased customer satisfaction and trust in the brand.
This wave of the fourth industrial revolution, having originated in Germany, has been quickly spreading to other countries with developed manufacturing sectors like the U.S., Japan, and China. Of course, South Korea is no exception. Led by the government initiative ‘Innovation in Manufacturing 3.0,’ the country’s manufacturing sector has been preparing diligently for the advent of a new age. The core of the initiative is the conversion of more than 10,000 factories, by 2020, into Smart Factories founded on the convergence of IT, Software, and IoT. This, in turn, is expected to lead to a countrywide innovation in manufacture that will help the country emerge as a new R&D hub of Northeast Asia. The Hyundai Motor Group has been a leader in this movement for the last few years; in 2015, it founded the Manufacturing Engineering R&D Center, a core focus of which has been to establish smart factories.
Smart Factory Technology: Manufacturing at its Apex
In June of 2017, the Hyundai Motor Group unveiled the Smart Tag System, composed of a position tracking sensor, high-capacity memory, and wireless communication chip. Under the system, each car on an assembly line is tagged with a small smart tag, so that the assembly robot can recognize the model being built and vary the assembly according to the model’s specification. It is quite common to have multiple models put together on a single assembly line during high-demand seasons, so the system allows seamless management of multiple models while keeping defect rates low.
Following the Smart Tag of the year prior, in 2018, the Hyundai Motor Group unveiled the automated inspection protocol for the Advanced Driver Assistance System (ADAS) mounted on vehicles. The technology, whose development started in 2015, employs six collaborating robots to inspect ADAS mechanisms in one integrated process. A significant improvement over the previous system during which many inspections took separate, time-consuming processes, the new process has the benefit of inspecting one car in as little as 85 efficient seconds. The technology is expected to be especially useful for production of self-driving cars, whose ADAS features are expected to be a lot more variegated than those of the current models.
In October 2018, the group unveiled the CEX (Chairless Exoskeleton), a knee-assistive wearable robot that helps a factory worker maintain a sitting position. The group subsequently released the VEX (Vest Exoskeleton) in September 2019.
Designed for the assembly line workers who work under passing car bodies and therefore must raise their arms at all times to do their work, the VEX is a wearable robot in the form of a lightweight vest—at merely 2.5 kgs. Its joint structure, which mimics that of a human, comes with the additional multilink muscle strength assistive device that alleviates musculoskeletal problems and improves general efficiency. The technology stands as a quintessential example of the rapidly automating Smart Factory’s modern identity: a place where humans and robots quite literally intersect.
The most notable advantage of these two wearable robots is in its versatility: that is, its fit, performance, and form are adjustable to a wide range of purposes and work environment. The group’s test application of VEX in Hyundai’s Alabama Plant and Kia’s Georgia Plant from January 2019 has been an undisputable success—the users noted that it provides exceptional muscular assistance without restricting user mobility. The group is planning to explore additional outlets for using the VEX, including other car manufacturers as well as other industrial sectors.
Smart technologies are not limited to manufacture: they are also found in product planning stages as well. The virtual vehicle design process unveiled by the group in December 2019 is one notable example. The process takes advantage of VR (Virtual Reality) technology, which allows the simulation of the model car and the driving environment in a virtual realm. The digital simulations then facilitate design/build checks as well as safety/quality assurance, providing the engineers with flexibility in making quick adjustments mid-development.
In March 2018, the Hyundai Motor Group opened the world’s largest VR design evaluation facility, capable of simultaneously hosting 20 people in a single virtual environment. The virtual development process then went into a test drive in June that year, and, soon found to be effective, became a full-fledged part of the group’s vehicle development process. At the VR evaluation facility, the engineers can view the virtual model inside and out as if it were a real model car, rotating it and/or changing the lighting as needed; even the functions on the virtual car work just the same way as real cars’ functions would. Many future car models of the group are scheduled to undergo the VR design evaluation process.
Developing a new model used to be costly, as doing so entailed the building of many test cars. These test cars, in many cases resembling the real one in both material and complexity, required much time and money to build. In the era of Smart Factory, however, virtual development has largely supplanted this costly and time-consuming process. VR development has been found to reduce development costs by 15% and development duration by 20%, but there is more: because the process allows quality assurance evaluation from the early goings of design, it has helped meaningfully from the quality perspective as well. The flexibility of the process—marked by the ease with which engineers can make changes—also allows the group to rapidly respond to the changing paradigms of the automotive industry as well as market demand.
Smart Factory Technology, as Shaped by AI
AI is widely considered the crux of the fourth industrial revolution, and as such, it has a mainstay presence in the Hyundai Motor Group’s Smart Factory innovations. In November 2018, the group founded the AIR Lab, an affiliate organization chiefly responsible for AI in cars. The AIR Lab has since evolved into a CIC (Company-in-Company) AIRS Company, now responsible for not only automotive AI but also mobility and life service areas. The AIR Lab within the company, though, is still responsible for the R&D for the AI-in-car manufacture area and its application.
Scanning & Deep-Learning of Paint Inspection Sheets is one of the lab’s noteworthy accomplishments. The technology, which uses AI to build a Big Data database of vehicle painting inspection, improves painting quality in the manufacturing process. Paint inspection sheets are manual records of painting quality made by the human inspectors, who write onto the sheet the codes for the flaw type and location. The lab believed that Deep learning, a branch of Machine Learning technology that allows AI to learn and approximate patterns of human-made data, could take advantage of this existing rich source of data.
Collaborating with the group’s Manufacturing Engineering R&D Center, the AIR Lab developed an algorithm that quickly scans and converts the paint inspection sheet’s information into a structured dataset including the time of inspection, vehicle model, type and location of flaw. The resulting Big Data database was deep-learned by the AI, during which it could reveal reiterating problems and model-specific issues that needed addressing.
Hyundai’s Ulsan Plant was chosen as the testing site for the algorithm from Sep. 2019 to Dec. 2019. The AI scanned an average of 400 sheets daily and showed a 95% scanning accuracy rate; as the system’s Big Data is in a relatively nascent stage, the accuracy rate is expected to increase as the database becomes larger. The engineers believe the system would lead to a meaningful decrease in painting defects. When it matures, it will be applied to the group’s other plants and production lines as well.
R&D efforts are ongoing to make AI useful in wheel alignment calibration as well. Wheel alignment, an angle adjustment of a vehicle’s suspension, is periodically required to keep the vehicle running precisely straight.
The researchers believe that AI can be employed during the wheel alignment process to calculate the ideal suspension angle and the corresponding nut torque setting. Here, machine learning comes into play again: the AI learns from the past wheel alignment data, predicting the ideal adjustment values for each vehicle model. These predicted values are instructed to the machine, which completes the alignment; the job results are then returned to the AI, which learns from its errors and seeks to reduce the error value, in a reiterating cycle of trial and error. As is the essence of Machine Learning, the more the system learns, the more precise it becomes; the engineers believe that a more mature online-learning based AI will be able to accurately predict the ideal adjustment values when given a new wheel alignment angle.
A Step Toward the Ultimate Smart Factory: Hyundai Mobility Global Innovation Center
The building of Hyundai Mobility Global Innovation Center in Singapore (HMGICs), whose construction in western Singapore’s Jurong Innovation District will be completed in 2022, embodies Hyundai’s visions to imagine and realize the ultimate Smart Factory. Thanks to the government’s aggressive support for technological innovation, Singapore has become a mecca of Southeast Asia and a home to some of the world’s best global companies, universities, and research institutes.
The Hyundai Motor Group plans to take advantage of this great local talent pool to develop an open innovation ecosystem in the HMGICs. That ecosystem will tangibly lead the group toward a diversity of future business models. The group also plans to employ the HMGICs as an important global location for developing new visions for future mobility.
To that end, the HMGICs will be in charge of the innovation project that encompasses all steps of car manufacturing—development, manufacture, and even sales—while serving the simultaneous role as a testing center for developing intelligent manufacturing platforms. A culmination of many futuristic technologies, such platforms will have IoT to gather real-time data, AI to analyze that data, and digital systems to instruct that data to production robots—a truly consummate, next-generation mode of manufacture.
In preparation for that ultimate goal, the Hyundai Motor Group is currently developing the technologies needed for AI-based, IoT-assisted, human-centered intelligent manufacturing platforms, whose testing grounds will be, at least at first, small-scale EV or customized production lines. But the HMGICs, the group believes, will eventually serve as a true spearhead of the group’s vision: becoming a global mobility solution provider, taking the literal leaps forward for humanity on the move.
Cutting-edge technologies can reduce the cost and time of development, and the profits from this can be relayed onto producing better cars; ICT technology can improve the lives of the workers, and not-so-far-away visions of mobility service are expected to transform the lives of the customers. The specifics may differ, but the principle is the same throughout. Behind the Hyundai Motor Group’s determined steps toward building Smart Factories is an insistent philosophy—to build better futures for humanity as a whole.