This article first appeared in Digital Edge, The Edge Malaysia Weekly on June 10, 2024 - June 16, 2024
Malaysia has made no secret of its ambition to be not only part of but also come out ahead in the Fourth Industrial Revolution (4IR), also known as Industry 4.0 or Industrial Revolution 4.0 (IR4.0). Over the past decade, the government has introduced a plethora of initiatives, with the latest being the National 4IR Policy and the Centre for the Fourth Industrial Revolution (C4IR), aimed at keeping the dream alive.
Yet, the leap into IR4.0 has not materialised as expected. A lack of capital and access to technologies, a dearth of skilled talent and stagnant salaries have impeded Malaysia’s IR4.0 readiness, particularly among micro, small and medium enterprises (MSMEs), which accounted for 97.4% of businesses in the country in 2022.
Malaysia’s IR4.0 journey began in 2015 with the National Internet of Things Strategic Framework. The framework did not translate into actual plans, which slowed things down. While the government launched the Industry4WRD Readiness Assessment Intervention Programme and the Malaysia National Policy on Industry 4.0 in 2018 to help build the momentum to IR4.0, it did not achieve widespread adoption.
Nonetheless, stakeholders believe that the advent of generative artificial intelligence (Gen AI) could serve as the great equaliser.
“IR4.0 is an important concept. It’s a way for companies to be a lot more global and a lot quicker [by utilising] technologies in order for them to be part of the global supply chain. [This is] to be able to be connected to their clients and suppliers to get the right information on time to do the right thing,” says Jaffri Ibrahim, CEO of Collaborative Research in Engineering, Science and Technology (CREST).
IR4.0 promises to be the next wave of digital transformation in the manufacturing sector, propelled by disruptive trends such as the surge in data and connectivity, advanced analytics, human-machine interaction and cutting-edge robotics.
“IR4.0 has always been talked about. We’ve had a number of years where the government has come together and stressed the point of IR4.0 and the next revolution. [However,] the adoption is not as complete as one thought it would be,” says Jaffri.
“There are successes. But as a whole, it hasn’t met expectations, even in the West. I think it’s no different in this part of the world.”
The disappointingly low rate of IR4.0 adoption has been exacerbated by limited access to capital, technological complexity and shortage of skilled personnel, notes Dr Afnizanfaizal Abdullah, head of innovation commercialisation at Malaysian Research Accelerator for Technology and Innovation (MRANTI). Understanding and navigating these new technologies can be daunting, especially for smaller enterprises without the resources of larger corporations.
But there is a need for the transition to IR4.0 to happen, especially among SMEs. As large companies are adopting IR4.0, this puts pressure on SMEs to remain competitive as suppliers or risk being left behind, says Jaffri.
That is because larger companies expect SMEs to be able to connect to their systems for data exchange and order processing. Moreover, these larger entities do not want to bear the cost of integrating with their suppliers. Instead, they expect real-time data on pricing and inventory.
“This way of doing business is happening right now with the bigger boys. If [SMEs] want to play at that level, [they] have to be connected to be able to be the supplier and partner of all these multinational [corporations]. If [they are] not, they will be too far down the line to be able to react and get a part of the business,” he says.
This makes it absolutely critical that Malaysian SMEs prepare for IR4.0, says Hanno Stegmann, managing director and partner at BCG X, the tech build and design unit of Boston Consulting Group.
Companies that continuously innovate to stay ahead of the game can ensure that they sustain their competitive edge. On the flip side, SMEs that do not adapt to new technologies, digitalise and leverage new tools are poised to fail or even go out of business in the long run.
This is where Gen AI comes in to disrupt the status quo. The tool that leverages advanced algorithms has democratised access to advanced technology for SMEs by reducing costs, simplifying usage and providing scalable and customisable solutions.
“Gen AI is a huge and significant accelerator for SMEs transitioning to IR4.0. Now, anyone can experiment with AI, thanks to the lower barrier to entry. What’s different with Gen AI is that it has shifted AI from the realm of experts to a broader audience through simple [and] conversational interfaces,” says Stegmann.
AI models are able to reproduce the knowledge embedded in their training data in real time during task execution, says K Raman, managing director of Microsoft Malaysia. Therefore, access to advanced skills are democratised.
“We see Gen AI as the great equaliser. In addition to delivering productivity gains, Gen AI lowers the barrier to entry into areas such as coding, graphic design and professional writing. For SMEs, this means a unique opportunity to leverage powerful AI capabilities to boost employee productivity, improve operational efficiencies, drive customer engagement and so much more,” he says.
The practical applications of Gen AI are vast and varied, says Stegmann. For instance, the technology can transform the customer service experience. Gen AI, with its ability to digest and analyse extensive datasets, can provide precise and real-time responses tailored to individual needs and preferences at scale.
Citing the example of an AI-driven personalisation engine that was developed as a result of a collaboration between Philippines-based GCash and BCG X, Stegmann says prior to the deployment of the engine, GCash could only manage about 30 customer relationship management (CRM) actions concurrently. The introduction of the AI-powered engine expanded its capability to more than 1,500 actions at any given moment.
“This engine can deliver over 225 million hyper-personalised messages daily, significantly boosting engagement and revenue by a factor of 10 compared to its previous CRM initiatives,” he says.
SMEs will be able to utilise Gen AI to streamline operations, enhance product development and optimise customer interactions, says Afnizanfaizal. Gen AI will be able to accelerate the transition to IR4.0 by automating complex tasks, enhancing data analysis and driving innovations such as AI-driven diagnostics and precision farming in sectors like healthcare and agriculture.
“The importance lies in staying relevant and competitive. Without transitioning [to IR4.0], SMEs risk falling behind in efficiency, market relevance and innovation capability,” he says.
“[This potentially leads] to reduced market share and diminished growth prospects. Transitioning enables them to harness the benefits of digital tools and technologies, [which are] essential to scaling and adapting to changing market demands.”
With Gen AI in the fold, SMEs can create and test digital twins of products without the high costs associated with physical prototypes, notes Jaffri.
Although some SMEs are enthusiastic about the potential of Gen AI to transform their businesses, others are cautious about the necessary investment and potential disruption to their businesses, says Afnizanfaizal.
These concerns stem from the costs involved in implementing AI technology, the complexity of the technology and worries about job displacement, says Jaffri.
“One of the primary worries is cost. Many SMEs operate on tight budgets and may find the initial investment in AI infrastructure and tools daunting … SMEs might also lack the in-house expertise required to deploy and manage AI solutions effectively,” he adds.
Nevertheless, Jaffri asserts that the long-term benefits of AI often outweigh the initial costs. To mitigate concerns about the complexity of AI implementation, partnerships with academic institutions and industry experts play a crucial role, he notes.
To some extent, this is already underway. SMEs have received a significant push towards Gen AI adoption, thanks to the substantial investments of major global corporations.
Microsoft, for example, has committed to invest US$2.2 billion over the next four years. This investment will go towards building cloud and AI infrastructure in Malaysia, creating AI skilling opportunities and establishing an AI Centre for Excellence.
These investments will lead to enhanced infrastructure, says Chin Chee Seong, national secretary-general of the SME Association of Malaysia. By having the data centres readily available, SMEs can tap into cutting-edge cloud infrastructure that eliminates the need for substantial upfront investment in hardware.
“The presence of these tech giants is expected to attract further foreign investments, creating a robust ecosystem for start-ups and SMEs … SMEs will be empowered to revamp their business models, enhance their competitiveness and successfully compete in the global market,” he says.
“[Furthermore] by expanding the pool of skilled workers [through training programmes to upskill the Malaysian workforce], the shortage of ICT (information and communications technology) talent in Malaysia [will be addressed] by providing SMEs with access to a larger and more capable workforce to drive their digital transformation efforts.”
MRANTI is supporting the segment by providing access to specialised facilities for technology testing and commercialisation, facilitating connections with investors and experts, and offering training programmes to upskill, says Afnizanfaizal.
“This support is crucial to effectively integrate and leverage AI technologies. Specifically, in our AI sandbox, companies are supported with technical facilitation, including technology experts from Nvidia and Amazon Web Services (AWS), potential investment from the funding partners and linking up with potential clients among public ministries and agencies,” he adds.
Collaborative ecosystems have to be established where government bodies, private companies and educational institutions work together to promote AI adoption among SMEs, says Jaffri. These partnerships facilitate knowledge sharing, resource pooling and joint development projects to make AI more accessible.
Ultimately, funding is crucial to enabling the transition.
For example, as part of the coordinated IR4.0 strategy, the Ministry of Investment, Trade and Industry (Miti) designated the Malaysian Investment Development Authority (Mida) to administer the Industry4WRD Intervention Fund, offering eligible SMEs grants of up to RM500,000 on a 70:30 matching basis. Companies can receive an upfront disbursement of 30% of the matching amount, with the remainder reimbursed based on eligible expenditures
Mida also provides Automation Capital Allowance (Automation CA), which serves as an incentive for manufacturing and service enterprises to boost productivity by investing in automated machinery and equipment. Under this initiative, the agency offers a 200% Automation CA on the initial RM4 million for expenses for labour-intensive industries and the first RM2 million for other sectors, including services. The Automation CA has been extended until 2027, with the quantum increased to RM10 million.
However, only 281 SMEs were approved to receive the financial support facility from 2020 to 2022, with a total grant value of RM101.4 million, Mida’s executive director of investment policy advocacy (manufacturing) Masni Muhammad told Bernama in October last year. Of the total, 82 SMEs were approved for the fund in 2020, 111 in 2021 and 88 in 2022.
To put this in context, there are more than 1.17 million MSMEs in Malaysia, which generated 38.4% of the country’s GDP and provided employment for 7.59 million people in 2022, according to the Department of Statistics Malaysia.
The poor uptake is mainly due to lack of awareness of the existence of such financial assistance in the first place, says SME Association of Malaysia’s Chin. The high cost to get the necessary verification and arduous requirements further complicate matters.
“Many SMEs are also not informed [about the existence of these incentives]. And if they are, many believe they will never get it due to the stringent and difficult application process,” he says.
For instance, for the Automation CA, applicants will need to go to Sirim Bhd — the government’s industrial research and technology organisation — for a verification of the automation equipment before submitting the application to Mida for approval.
“[Previously,] some SMEs may have been reluctant to apply because there were some application costs. But with the increase in quantum, they should be more willing to apply now,” says Chin, adding that there needs to be more promotion of the initiatives provided to create awareness among SMEs.
This is because with financial support, companies can invest in cutting-edge technologies, upgrade their infrastructure and access the expertise needed to innovate and stay competitive, says Jaffri.
An example is the CREST R&D grant, which provided RM130,000 to TT Vision Holdings Bhd and Universiti Sains Malaysia as part of an industry-academia collaboration. “This funding led to the successful development of an automated online inspection system for solar wafer micro-crack detection, leveraging photoluminescence imaging and a machine learning-based algorithm,” he says.
Meanwhile, MRANTI facilitates access to funding through initiatives such as the National Technology and Innovation Sandbox (NTIS), which helps companies to deploy their innovations within a safe and controlled environment.
Every bed of roses has its thorns. While AI comes with a host of benefits, there are issues to be addressed.
Models used by local SMEs need to be appropriate for local contexts to derive meaningful benefits from Gen AI, says Dr Rachel Gong, deputy director of research at Khazanah Research Institute.
“This means using training data based on Malaysian markets, practices and supply chains. This is where the rubber meets the road. There needs to be enough information shared about all these localisations that can then go into training models that suit local SMEs,” she says.
“Historically, data has been a very big challenge and I don’t expect that to change quickly. Language is another challenge for Gen AI models in Malaysia as the models are not equipped to sound natural in Malaysia’s multilingual and multicultural society.”
Moreover, it is essential for policies and regulations to support the transition for SMEs to IR4.0, says MRANTI’s Afnizanfaizal. These are ethical AI policies that allow governance of data as well as the alignment of existing policy and guidelines in various industries.
“On data security and privacy, the first step would be improving data regulations to take into account how data fed into Gen AI tools are governed. At this point, we have data regulations for personal data, but they only cover transactions, which would presumably exclude data used in Gen AI training,” says Gong.
“Next, there would have to be some sort of grievance process by which consumers and businesses who feel their data might have been leaked or misused by AI can seek redress. Such a process may mean having to retrain a model so developers and deployers need to be consulted.”
As Gen AI is heavily dependent on large amounts of data, it is important to get support from various agencies and ministries to assess certain data for training the models with better accuracy, says Afnizanfaizal. MRANTI will focus on promoting localised large language models, in which models are trained with local data rather than using existing data in public repositories.
“For policies to be effective, they must be developed in collaboration with recognised experts in the field. In Malaysia, it is important to focus on policies that support upskilling and training as IR4.0 necessitates a workforce that is agile, skilled and capable of adapting to new tools and methodologies,” says BCG X’s Stegmann. In any case, there needs to be a deep understanding of Gen AI’s potential and limitations across all levels of an organisation, he adds.
BCG advocates for a strategic approach via the 10-20-70 formula. This means 10% of the effort will go to developing an adequate machine learning model, 20% of the effort is focused on sourcing high-quality data and implementing the technology, and the lion’s share at 70% is dedicated to change management. This is to develop new business processes and transform existing operations.
“There’s a common understanding that transitioning to IR4.0 and embracing technologies like Gen AI isn’t an option — it’s a necessity. However, there’s also a palpable sense of apprehension,” says Stegmann.
“It’s akin to a rabbit frozen in the glare of an oncoming snake. IR4.0 isn’t the snake, though. It’s the ladder to unprecedented opportunities.”
Save by subscribing to us for your print and/or digital copy.
P/S: The Edge is also available on Apple's App Store and Android's Google Play.