The ongoing semiconductor shortage has been a topic of conversation and debate the world over. It is a phenomenon which has affected multiple industries and tens of thousands of companies. The COVID-19 pandemic is considered the main culprit behind the ongoing troubles in the semiconductor supply and availability. However, if examined closely, it becomes evident that while the pandemic played a major role in the current scenario, the reasons why some industries are suffering more than others are deeper and related both to the customers and manufacturers.
It’s not that the demand isn’t there for semi products. Design News points out the following drivers:
- It’s projected more than 75 billion IoT devices will be installed worldwide by 2025
- A ‘smart car’ will generate more than 4,000 GB of data/day
- In the period of 2016-2025, the volume of data generated will increase tenfold to 163 zettabytes
- 5G technology will reach 1.3 billion subscriptions by 2023
- Emerging ‘smart’ technologies (tablets, drones, smart phones) will reach $6.9 trillion by 2030
Even before the pandemic hit, the semiconductor industry, while being at the leading edge of product research and development, needed an overhaul in terms of adopting better automation, IoT and other Industry 4.0 technologies. The cost of a semi plant—between $2B to $4B—and the time it takes to build—2 to 4 years—means many semiconductor companies are running on older, close to obsolete equipment and systems. Companies in other industries, such as automotive, oil and gas and food and beverage, had already embarked on the transformation journey, understanding the benefits it had to offer and witnessed the proven success of digital transformation.
The pandemic came in as a wakeup call. There were two waves of ‘whiplash’ for the semiconductor industry. The first came with cancelled orders from primary OEMs, a result of the pandemic shut down: car sales were down, elective surgeries were cancelled, stores were closed. The second wave brought recovery—and in a flash, demand was up over pre-Covid rates. As an article in SIA states: “during the second quarter of 2020, when automakers understandably reduced production, chipmakers saw surging demand for semiconductors used to enable remote healthcare, work-at-home, and virtual learning, which were needed during the pandemic.”
For semiconductor manufacturers, this shift in demand meant urgent rethinking of supply and production strategy, and a shift in focus towards high performance modern products, in demand from the medical, electronics and gaming industries.
2021 proves to provide dramatic recovery, and semiconductor manufacturers are now faced with buoyant demands across the board; the forecast is for 10% or more revenue growth for companies building chips.
The current industry transformation
Deloitte identified four primary characteristics pertaining to transformation in the semiconductor industry:
1. Disrupted Digital Transformation Plans
Almost half of the survey respondents believe their digital transformation strategy has to be modified to adapt to market changes, strategic objectives and technology. It begs the question whether flexibility and adaptability were built into their strategy at the very beginning? The answer most probably would be a resounding ‘No’.
Semiconductor manufacturers faced massive variations in demand during the course of the pandemic, which meant faster introduction of new products, faster experiments, shorter R&D cycles and the implementation of automation, control and data analytics across the operation. All this change, unless anticipated, slows down existing transformation efforts and brings management back to the drawing board. Digital transformation if pursued in a piecemeal manner will always create results which are sub-par, and when faced with disruptions which affect the entire operation, areas which haven’t yet been prioritized for transformation may become bottlenecks.
A key takeaway is to look at digital transformation as an ongoing, all-encompassing exercise, which begins with a solid foundation: the MES platform. The choice of a MES which controls and orchestrates a given plant can be the difference between an agile and a rigid operation, which in turn affects the go to market time, operational flexibility and overall throughput of the plant. Selecting a MES which is industry specific, delivers Industry 4.0 tech, and allows gradual but continuous scaling and automation should be the very first step towards a sustainable and rewarding transformation effort.
2. A Need to Look at New Frontiers
The way in which most semiconductor manufacturers conduct business is quite straightforward: the sale of chips ordered. However, chip manufacturers are looking at more than just sale of chips and actively incorporating edge computing, AI, 5-G and IoT related features and services in their product mix.
This need to go beyond the manufacture of the chip itself. Providing services typically associated with other industries requires the establishment of partnerships and other collaborations. In turn, it requires more than just deals to be made; it requires operational re-engineering and connectivity across partner companies within a given value chain. While chips provide the computing power, with the value addition of AI and/or IoT or other services being provided at the manufacturers’ level, it opens up new avenues for revenue generation.
Here again, the MES proves essential, as the application not only covers every single aspect of the manufacturing process, but allows for data to flow seamlessly across applications and even organizations. This integration is essential for a typical manufacturing business to incorporate service-related aspects within their product portfolio. Data from manufacturing and experiments contributes directly towards improvements in a particular chip design. When collaboration and feedback loops are established across the supply chain, technology partners can play an active role in product development. When additional services are incorporated, the MES acts as the overarching application, which enables the technology bundle to be upgraded, from both a collaboration and engineering perspective.
3. Modified business models and new solutions
Chip manufacturers need to look at their products from the perspective of software vendors, where the sale of chip is just the beginning of a transaction and not the end. Semiconductor chips perform computing, and data flows through them for every single transaction which happens on any given device they are installed on.
Data is the currency of Industry 4.0 and thereby should be monetized by chip manufacturers. AI resides on chips and manufacturers can leverage this to go beyond the simple chip manufacture and sales cycle. The service provided can be to analyze and present the data which passes through the chip and help customers improve their products.
The MES platform is the application which brings AI and ML functionally into semiconductor manufacturing. MES also integrates with essential point and automation applications and process equipment to capture shop floor data. Through IoT and edge computing, MES triggers immediate, automated actions to real-time shop-floor situations. AI/ML works in the background, continuously learning from current shop floor trends and past data to establish multiple scenarios, which control and orchestrate everything from material logistics to recipe management, constantly optimizing workflow and prompting actions based on data which changes dynamically.
For semiconductor manufacturers eying the ‘X’ as a service model, the need to have a clear understanding of how AI/ML can be implemented within their products to benefit customers and improve their own revenue streams, will often begin with their own transformation efforts with the MES at its center.
4. Digital All the Way
Chip manufacturers have always focused on designing and building better chips following Moore’s law, keeping a stringent focus on miniaturizing and improving core products. The Deloitte survey however indicates a digital proliferation in the industry, which goes beyond manufacturing better, faster, smaller and smarter chips; it points towards the addition of software capabilities and additional technological features. This helps customers build around the original product, modifying its use to innovate and create better product offerings of their own.
All this begins when transformation is pursued by the chip manufacturers as ‘the strategy’, when AI, IoT and edge computing, along with hyper automation are pursued not as a mere goal, but as the very future of the operation. This complete digitalization happens only with the MES. It is the application which brings flexibility, scalability and Industry 4.0 technology into semiconductor manufacturing.
The MES platform plays a critical role in digital transformation, which needs to be incorporated into the company DNA. It brings manufacturing excellence to the forefront.
The only consideration which the semiconductor manufacturers need to make is which MES to use as the foundation of their transformation? We suggest a MES which makes the transformation and resultant transition smooth and seamless, which is able to connect with needed point and legacy applications, while enabling better automation, orchestration and control. An MES which brings AI into the solution set. It should incorporate IoT through its functionality, enabling faster throughput, better experiments management and wafer level visibility. The application also needs to be technology forward, allowing graduation from status quo towards full automation without disruption, which means that the MES vendor should have experts on board which understand the nuances and challenges that the semi industry brings.
The future of digital transformation in the semiconductor industry is to pursue industry 4.0 to go beyond chips. It begins with a re-engineered operation and a re-imagined business, which at its core needs a MES capable of supporting this large scale and all-pervasive strategy. Disruptions like the pandemic and a ‘status quo’ operating philosophy will always be a threat to manufacturers in the semiconductor industry. Those who adapt and incorporate agility, resilience and digital into their operation will be the ones who thrive.
Download our new eBook and Technical Guide that have everything you need to know about the advanced capabilities that an modern MES should support: