There is an increased realisation that companies, particularly in the manufacturing sector, can derive significant business value from the use of data analytics, artificial intelligence (AI) and machine learning (ML) in their digitalisation efforts.
A major point brought out during an industry roundtable on, Transforming Manufacturing by leveraging value from AI, ML and data analytics, organised by iTnews Asia, in partnership with AWS on May 4, 2023, was that digital transformation in the traditional sense was not working.
According to a study, while 68 percent of manufacturing companies have started their smart factory initiative within the past year, only 30 percent have achieved success or are on the verge of achieving successful business value.
The more telling statistics is that more than 40 percent of the manufacturers who embarked on their Industry 4.0 journey are stuck in what is called “pilot purgatory” as they are not able to scale to production to ensure meaningful business impact.
AWS’ ASEAN Head of Technology for Strategic Industries and Digital Manufacturing Expert, Dr Daniel Kearney, noted that the vast majority of companies struggle to capture the full potential of their digital manufacturing transformation initiatives due to a lack of alignment to scalable business outcomes and a scarcity of internal skills to expand the proof of concepts into production-grade solutions.
Setting the tone for an engaging discussion, Dr Kearney gave a brief background of AWS’ digital manufacturing capabilities and how AWS supported Amazon.com to evolve its data-driven operations and expertise to where it is today.
By catering for the huge infrastructure needs of Amazon, AWS has developed key insights about what works in the manufacturing sector and the company has broken down that knowledge into several key solution areas within the manufacturing vertical, he said.
Dr Kearney highlighted the role a data-connected manufacturing ecosystem can play in reimaging the end customer experience; two areas that are not normally considered together.
“We are hearing from our customers how connecting smart products back to the engineering and design workflows supports immediate product improvement and enhanced customer experience,” Dr Kearney said.
This visibility from the real-life use cases allows AWS engineers to do higher fidelity design and simulation, the number of prototyping steps and ultimately reduce time to market for innovation. This in turn helps in moving to actual smart manufacturing internally within the factory by automating processes and quality controls, thus improving productivity and safety, he said.
Sustainability
Sustainability was another key solution area and was linked to driving improvements in cost by driving down carbon usage, Dr Kearney said.
Related to this, another key focus area for AWS is to double down in its logistics and supply chain around energy usage. This allowed the company to monitor and optimise assets and do the fundamentals right, Dr Kearney added.
AWS has made its proprietary supply chain forecasting models available to the public and “we’ve driven this further with more integration with partners”, Dr Kearney said.
He noted one of the reasons many companies who embark on their smart factory journey are stuck in “pilot purgatory” was that they usually have multiple plants or facilities.
The common pitfalls are that these facilities have people with their own particular needs and use cases, he said.
“And they will naturally prioritise based on what they want and what they need for their facilities. And then we'll start to have individual deployments where the return on investment (RoI) starts to get localised, with bespoke point solutions in each of these facilities,” Dr Kearney said.
He noted that while these solutions can be very useful in individual use cases it is very difficult to scale them companywide.
“Every single new use case is a rebuild and capability doesn't scale up into your organisation to the enterprise level. So, then we start to have this limited repeatability and sub-optimised processes,” he said.
Dr Kearney added that despite best intentions, the enterprise IT-OT (Operational Technology) divide persists. “So, the data silos within our organisation continue because everybody has the data that they need for their own localised particular solution,” he noted.
Lack of focus
Expanding on Dr Kearney’s “pilot purgatory” comment, Siemens Digital Industries Software’s APAC pre-sales director, Luc Van Laere, said many companies get stuck because they often do not have the focus on what they want to achieve.
“When you have an initiative for moving to Industry 4.0, you really need to be clear on the scope you want to cover. Don't try to solve world hunger, and change your entire company,” Van Laere said.
Companies need to understand why they need this and start relatively small, while not forgetting what the end goal is, he said.
“I know we're talking about manufacturing, but depending on what you do, the story is not limited to manufacturing, irrespective of the industry.
“It starts with the design of the product, and from all the ideation, the design, then you move to manufacturing; and then you move to your customer. And for every customer your focus point is going to be different,” he said
Accenture’s Engineering and Manufacturing Lead (South East Asia) - Industry X, Sathvik Rao, noted that as consultants “we (Accenture) get involved in a very early stage” in developing a blueprint for companies seeking to develop an Industry 4.0 strategy.
“Clients often come with simple objectives, such as I want to double my volume by the next two years”, Rao said.
The problem starts when we try to dissect that and get a solution that ushers in a manufacturing transformation, he added.
“The challenge that we face is every project starts with a pilot. We need to understand what problem statement they are trying to solve.
“What is there to add to your business case? What are the cost and revenue drivers that will get the business case worth the investment that they are committing?
“So that is one big challenge that we see, starts with the technology rather than cutting to the problem statement of the client,” Rao said.
The other challenge is that, typically, in these kinds of projects, there are multiple stakeholders, one is your business, which is the head of manufacturing, then you have your supply chain, and then you have a design team, he added.
So how do you manage all these stakeholders and satisfy their needs and get to a successful pilot and then the actual project? “That's another added challenge that we see”, Rao said.
What to look for in a technology partner
To a question about how companies, particularly small and medium-sized enterprises (SMEs) should choose their technology partners for their digital transformation journey, Dr Kearney noted that company size was really not an issue; it was more about the speed at which a company can leverage technology to innovate and realise the business benefits that was a “great equaliser”.
Dr Kearney noted that as a part of his role he “speaks to many companies” that are challenging the “largest companies in the world with the use of technology”.
“I think technology is the great equaliser. We have seen companies use technology to become extremely fast and agile and improve efficiency,” Dr Kearney said.
Talking about Amazon’s own journey he noted that the company has grown to a level where it really needs to be cognisant of the environment around it.
“We need to internally work on maintaining the speed and how we actually continue to be super innovative and at the same time don’t lose all the attributes of a large organisation by using our digital capabilities”, Dr Kearney said.
He, however, noted that even with the best-of-breed technology, developed in-house, AWS still struggles with speed of innovation, because of the scale of things. “And we are working very, very hard with that”.
On the specific point of what companies should do to further their digitalisation journey, Dr Kearney noted that there is a need to maintain the fundamental principles within the organisation.
Technology is only partly the answer, Dr Kearney said, adding that it's the organisation's appetite for innovation in terms of people and the teams that they set up to bring in agile processes that are important.
“Don’t punish failure, teams and people will make mistakes but they should be rewarded for continuous development and learning quickly from failures”, Dr Kearney said.
“When using technologies, the key thing is to keep a big organisation acting like a small one. And of course, the small ones have the advantage of having at their fingertips large (technology) platforms to enhance their influence and scale,” he added.
Cloud as a technology enabler
Dr Kearney noted that a cloud technology partner can deliver, especially to SMEs, technology which previously was only available only to large corporations.
Adding to Dr Kearney’s point, Siemens’ Van Laere said: “Take the example of a company called ZipLine, it's a company making drones for delivering medical supplies in Africa.
“They're a very small startup and, of course, the use of drones is a high-tech endeavour. You need all the technology you can think of, including CFD (computational fluid dynamics) analysis and that’s one of our products, ‘How does the (air) flow go around your drone? And is it stable in the air?’
“This is normally the job of supercomputers. Now we (Siemens) just deliver that on AWS as a service. And these companies take three months, or what have you, to do their calculations,” he added.
Accenture’s Rao noted that while his company didn’t “really work with SMEs specifically”, the subject was “close to his heart”.
“As a technology partner, we enable large companies to become nimble and agile. Larger companies come to us and ask ‘How can we compete with the smaller players, platform players that are coming from nowhere and disrupting our business’”, Rao noted.
From that perspective, Rao said, “I think the cloud service providers, hyperscalers like AWS, have democratised technology”.
They have given SMEs access to technologies that previously were only available to large enterprise customers, Rao said and added that a “one compartment company” can go and subscribe on the cloud.
“One best example is the AWS monitoring tool, which for, I think, around US$7, (S$9.40), you can get a subscription where it comes with four devices and a gateway which you can deploy and monitor machine conditions.
“I think even predictive analytics has been added recently to that,” Rao said.
I remember 10 years ago only large engineering companies had the budget and ability to do such kind of projects, he added.
“Now you can do it at US$700 a year subscription… that is the way hyperscalers have democratised technology,” Rao said.
He gave another example of Siemens.
“I was at their partner event a couple of months ago. There was a very good example from an electric motorbike company from India, where the founders recounted how, as a 10-member team sitting in a garage, they started an electric vehicle company.
“They were able to leverage all the capabilities that Siemens Teamcenter PLM (product lifecycle management) offers to a large enterprise using a subscription; minus that they would have not been able to deploy on-premise servers that would be required to get the same functionality,” he said.
Using a combination of AWS, Siemens PLM and Teamcenter on the cloud in a subscription mode, from day one of their incubation, they were able to start using the same enterprise software as large companies use to drive innovation, Rao said
Van Laere added that the training aspect was also “extremely important”.
As technology providers, we need to not just deliver the technology and software but also work on the “cultural change within a company”, he said.
