Cyber Physical systems and Digital Twins ... Misconceptions and realities

Ever since the industrial revolution, the manufacturing industry evolution has always been driven by new technologies that facilitate improvements such as increased throughput, lower costs and reduced downtime.

In modern times, the evolution is with the confluence of technologies integrating cyber-virtual and cyber-physical systems, culminating in I4.0. Cyber–physical systems (CPS) and Digital twins (DTs), the new paradigms of virtual-physical-cyber integration are the key enabling systems to manage transformation of product lifecycle reduction, product mass customization, and growth in product variety, reduction of time to market.

CPS and DTs share the same essential concepts of an intensive cyber–physical connection, real-time interaction, organization integration, and in-depth collaboration. Why then, do these two concepts exist in parallel, and why are they used in different fields? What are their respective core elements? And which is more suitable in practice? These are some of the moot questions, often heard in forums.

Different perceptions, some misconceptions and many unanswered questions has drawn me to deep dive and try uncover the differences and correlation between Cyber Physical Systems and Digital Twins.

CPS emphasize on sensors and actuators with the concept of system thinking, integrating computational and physical processes from networked products and operations. It defines how a physical system integrates sensor, communication, computing and control in a large scale cyber infrastructure to provide real-time sensing, information feedback and dynamic control for complex systems, making the mapping between the cyber and physical worlds one-to-many.

CPS are more foundational, as they do not directly have reference to implementation approaches or particular applications. Efficient collaboration between the cyber and physical worlds enable the functions of CPS.

DTs emphasize on models and data with the concept of using a digital copy of a physical system and provides a complete digital footprint of product. Using bi-directional dynamic mapping, the physical entities and virtual models co-evolve and have a similar appearance, like twins, and the same behaviors, like a mirror image. The mapping relationship between the physical and digital worlds helps provide a one-to-one correspondence to directly optimize the operations and adjust physical process through feedback.

Digital Twin system has multiple definitions and each has a different perspective. From product designer perspective, product life cycle management (PLM) is the key platform to use product design model to monitor, test, control, and service the products in the field. From a user perspective, user data from the usage is the key source to model the purpose (quality, efficiency, maintenance, etc.).

In Summary both CPS and DTs pave the way for smart manufacturing by forming a closed loop between the cyber/digital and physical worlds based on state sensing, real-time analysis, scientific decision-making, and precise execution. However, by virtue of its virtual models, a DT provides a more intuitive and effective means of improvement in engineering.

Through continuous data integration, the DTs’ capability of offering related solutions can be strengthened. Virtual models can be used as a supplement to enrich the composition and functions of CPS, so DT technology can be considered a necessary foundation for building CPS and for opening the way to the realization of CPS. The combination of CPS and DTs would help manufacturers achieve more precise, better, and more efficient management.

The Shifts of Technology… that Tests the Reliability of Systems

When I had to make a choice of Flat screen CRT TV in the early 2000s, Sony Trinitron, which was then known in the market as the emblem of premier CRT televisions, was my select. Following years, slimmer, better-looking HD-ready LCD and LED TVs came along…but with the CRT TV meeting the requirements, I was more than happy to pull along for many more years. It is a different matter that eventually, my family’s choice took precedence over mine and I had to upgrade to a sleeker, wider, digital, smart TV. However not prepared to junk the impeccably performing CRT TV, I continue to use the bulky TV from the sunset years of its technological life; waiting for it to have a natural death.

Needing a convincing, why my CRT TV hasn’t quite died out yet, it takes me to the question of understanding reliability.

When reliability is the probability of performing a required function without failure for a stated period of time. The moot question is on deciding on the the goals for the service life, design life and market life characterized by their varying degree of independence and interdependence. While the service life delineates wear characteristics of the product, design life defines the life expectancy of the product and the market life outlines the time the product continues to be sold in the market.

The unparalleled strength of Sony brand that has all along been my choice has endured for decades with exceptional reliability which allowed it to charge a significant premium for its products. However, in this era of rapid rate of technological advances where the time between new product roll outs shrinking and the average lifetime of products falling; cost pressure and the time to capture Mind share takes precedence. As a result, the design philosophy has to cope with the shifts of technology…that would test the reliability of systems.

This would entail a philosophy, where technology obsolescence precedes the life expectancy of a product; while at the same time the design margins are optimized to make the best of market life. To make this work, knowledge on Technology obsolescence and Physics of Failure (PoF) are both important. Accurate prediction of technology obsolescence being near to impossible, the Tech industry is adapting to this idea of planned obsolescence of product technology and integrating physics of failure approach to the design process for optimizing the design margins.

The PoF approach incorporates reliability into the design process by establishing a scientific basis for evaluating materials, components and mechanisms. PoF models are then applied to Highly Accelerated Life Tests (HALT) of physical prototypes pushing the product beyond its operating stress conditions to accelerate failure. HALT with multi-sensor data analysis helps determine the operating and destruct limits of a design and with a prognostic intelligence model the faulted components Remaining Useful Life (RUL) gets defined to either strengthen or optimize the design margins.

In this era of tectonic shifts in technology, where the pace at which products arrive and fade from the market is fast shrinking, the reliabilty of systems are being put to severe test. Advancement of Digital Technology in Simulation, Test, Measurement, Control; powered by Artificial Intelligence is now fast gearing up to meet the new set challenges.

The Factory of the Future…Economies of Scale to Scope Economies

Factory systems has come a long way from the process of product replication beyond the "craft" stage -- where highly skilled craftsman treated each job and each product as unique -- to a more efficient mode of operations aimed at economies of scale.

Rather than making products one at a time, new efficiencies came through standardized and interchangeable parts, fixed and simplified production procedures, use of less skilled workers with divisions of labor, mechanization and stricter process and quality. Ford’s assembly line considered revolutionary for mass manufacturing was one of the first where the work moved to the worker through conveyers. This sharply reduced the time it took to build a car and with this innovation Ford’s Model T turned out to be an automobile for the multitudes.

With the introduction of automated assembly lines, modern factories emerged and firms introduced a series of product and process innovations that made possible the efficient replication of a limited number of designs in large quantities. These facilities captured economies of scale by standardizing and integrating production processes in large, centralized facilities; closely coordinating the flow of each process; dividing and specializing labor; automating tasks; and imposing rigid managerial, accounting, and other types of controls.

But while large scale factory organizations do provide higher worker and capital productivity, the rigidity in processes and equipment's is seen as a showstopper to introduce new products or processes quickly and economically to meet the demands of customers with distinctive tastes. The end of the era of “Model T”, considered one of the most successful models in the Automotive industry is a typical case of how changing tastes can beat out standardized offerings even on low cost and how the producer- Ford’s efficiency crusade with standardization, where buyers couldn’t even have a choice of color had to face bankruptcy for pushing process rationalization and scale of economies too far.

With consumers’ choice progressively shifting from standardized, mass-produced products, in favor of custom-made solutions that fit specific needs, the craft approach of "flexible specialization" – that rely on relatively skilled operation – is seen to supplant mass production. Going by the consumer trend, would the once dominating conventional mass production get more or less obsolete? Would the craft mode of production resurface? Would there be a tradeoff of flexibility over efficiency?

Well, The Factory of the Future (#FoF) is here and it assures best of both worlds!

The factory of the future has three common overarching characteristics: connected, automated and flexible digital shop floor processes; new relationships between operators and machines; and the structure, location and scale of the factory.

Shift from scale economies to scope economies - New operating concepts and engineering techniques, along with advanced computerized technologies, would reduce the tradeoff of flexibility for efficiency with a shift from scale economies, as in conventional mass production, to scope economies -- efficiencies gained in the design and production of multiple products.

Merger of workflow and machines - There would be a complete consolidation of devices and process management. "Process and device" will be inseparable; physical things will become part of the process. The work flow would cease to exist as an independent logistical layer; it will be integrated into the hardware.

Batch-level to unit-level visibility - Each of the physical assets, semi-finished products, and even component parts will be equipped with a digital identity that would describe the exact location and condition of fixed and moving assets in real time ubiquitously throughout the supply chain.

Robots to Cobots - Emergence of IoT technologies and advancement in Artificial intelligence would change existing automation and there would be an evolution of Robotic automation from synchronized operation, simultaneous operation to assisted cooperation with human. Wearable technologies would digitize the workforce that would assist human-robot collaboration for robots and operators to co-operate closely to handle and process products jointly with both the agility and reliability offered by robots and the flexibility offered by human operators.

Fixed assembly line to flexible virtual assembly line - The rigidity of fixed conveyorised assembly lines would give way to the notion of a portable assembly line that would replace conveyorised lines with Autonomous vehicles that would move the work-in-progress from station to station with the flexibility to reroute the assembly process on a moment's notice.

Large, centralized facilities to local micro factories - The need for speedy delivery, short cycle times, and the flexibility to respond to uncertain demand and product customization requirements, would pave the way for micro factories located at the local marketplace, where proximity to consumers is a key value driver.

How it would work?

In the #FoF scenario, raw material will say, "I am the block that will be made into product X for customer Y." In an extreme vision, this unfinished material already knows for which customer it is intended and carries with it all the information about where and when it will be processed. Once the material is in the machine, the material itself records any deviations from the standard process, determines when it's "done".

Robotic autonomous vehicles will allow Manufacturing Execution Systems (MES) to route the virtual assembly line on a moment's notice to reflect the real-time demand gleaned from a company's Enterprise Resource Planning (ERP) tool. This will allow adaptive factories to produce exactly what the market demands based on sales data.

Whatever the future may hold for factory automation, technology will continue to build on centuries of amazing human innovation and ingenuity. The factory of tomorrow will no doubt be more capable and adaptable while yielding cost effective, highly customized products to meet the demands of customers the world over.

In the era, when time between new product rollouts shrinking and average lifetime of products falling, an observed Market opportunity or New technology...what is the pivotal factor for business success? You may want to read my post Market Pull and Technology Push…Can they be two sides of the same coin?

Will new business models shift from products to services? You may want to read my post Creating business model...where digital meets physical

Creating business model...where digital meets physical

With the financial year closure fast approaching, it is the decisive time of the year to sprint and pace to the finish line. It is also the time to re-visit strategies and re-define short term and long term business plans.

Leading a machine building company that is in the capital goods industry, I am now in the midst of reshaping our strategy and defining the business plan for the next year and the years after.

The capital goods industry largely operates on a product focussed ownership business model where the manufacturers like my organisation have traditional ‘tangible’ production costs and the revenue is mainly generated from the product sale. The Service revenue is limited to break-fix repair, maintenance contract.

As the saying goes Customer—not the competition—is the key to a company's business strategy, I had in the past built our strategy around the belief that focusing on competition leads the strategy inevitably to the notion of sustainable competitive advantage, while focusing on the customer leads to the notion of value.

Now when I am drawing out our strategy for the future, I realise that pressure is mounting on manufacturers like us to look downstream to uncover new value creation opportunities by helping customers use their products to meet specific outcomes. I recognise that large-scale shift from selling products or services to selling measurable outcomes is a significant change that will redefine the base of competition and industry structures.

This would require companies like us to forge new ecosystem partnerships centred on customer needs rather than individual products or services. This evolution will call for reshaping the operating model and accept customer’s investment model of Opex from Capex.

This transformation path from a product-oriented strategy to a combined product-service strategy remains a complex concept with the major obstacle being monitoring mechanism for product usage and related data, financing options...

While there are too many questions that come my way to transform the strategy for the immediate, it will not be too long before this turns in to reality.

In the upcoming years, the Internet of Things will dramatically alter the vision of tomorrow’s manufacturing, already declared to become the “Fourth Industrial Revolution”. This paradigm shift will involve symbiotic networks of people, smart materials, smart devices, smart machines and Smart factories that will form net-centric societies or cyber-physical social systems.

This latest wave of technological change will bring unprecedented opportunities to business and society. It will combine the global reach of the Internet with a new ability to directly control the physical world, including the machines, factories and infrastructure that define the modern landscape.

With the ability to monitor machines that are in use at customer sites, makers of industrial equipment like us can shift from selling capital goods to selling their products as services. Sensor data will tell the manufacturers like us how much the machinery is used, enabling the manufacturer to charge by usage. Service and maintenance would eventually be bundled into the hourly rate, or all services would be provided under an annual contract.

Emergence of Industrial IoT would turn products into services, and in order to fully capitalize on that fact, it would call for manufacturers like us to make some fundamental, systemic changes in the way the businesses are conducted and strategies are drawn. Advent of Industrial Internet will be transformative and it will change the basis of competition, redraw industry boundaries and create a new wave of disruptive companies. Disruption will come from new value creation made possible by massive volumes of data from connected products, and the increased ability to make automated decisions and take actions in real time.

As the Industrial Internet gains broader adoption, businesses will shift from products to outcome-based services, where businesses compete on their ability to deliver measurable results to customers. Such outcomes would range from guaranteed machine uptimes on factory floors and the qualitative results. Accordingly predictive maintenance and remote asset management one of the most widely cited application of the Industrial Internet will reduce equipment failures or unexpected downtime based on the operational data now available.

My future ready strategy will require new levels of collaboration across an ecosystem of business partners, Software platforms that would better facilitate data capture, aggregation and exchange across the ecosystem. This sure would bring together players that combine their products and services to meet customer needs.

I realise that the shift from products to services to outcomes will not only disrupt internal operations, but will also impact how we go to market. With the access and control points more open and fluid in a digital marketplace than in traditional markets, my strategy for future will have to factor competition from a broader set of players, including digital pure plays founded on new models and platforms from their very inception.

In summary I realise that if we don’t transform the way we do business, it is like waiting for the expiry date. It’s not just about changing the way we do technology... but also changing the way we do business.

Market Pull and Technology Push…Can they be two sides of the same coin?

It is believed that market moves at the speed of culture and this signifies the pace at which products arrive and fade from the market. This speed is hastened by new technology and the generation of consumers using it the most. Millennials have come of age during a time of technological change, globalization and economic disruption.

In this era of technological change, where the time between new product rollouts is shrinking, the average lifetime of products has also been falling. The pace of innovation has ramped up to ludicrous speed and the products are coming to market at a breakneck speed. Manufacturers are continually outdueling one another for Technological Newness and are competing to be the first to capture the Mindshare of customers.

Where speed to mindshare dictates success or failure. Companies that take too long to commercialize their products may fail to capitalize on a narrow window of opportunity before competitors swoop in and pass them by. Companies also look at maximizing the product life-cycle and the peak sales increase overall aggregate sales by leveraging an early entry in to the market. Consumer facing industries with a high concentration of innovation and where time to mindshare are the critical value drivers, the trend is an indication to go by.

In the early days of basic phones, average lifespan of a mobile phone was significantly longer: three years and more and now the average time smartphones spend on the market is just six to nine months. To cope with the shrinking product life span, the development cycles of smart phones are fast shrinking and now believed to have dropped down to 90 days and less.

The key to this possibility in the electronics architecture is the availability of SoCs (System-On-Chip ASICs), where hardware dependency in a smartphone have been ephemeralized – converted into information complexity inside the SoC. The combination of commodity SoCs, ephemeralization and the rise of 3D printing has cut time to prototype much further.

Not long ago, Auto makers shrinking the amount of time to bring a new product to market was unthinkable. Over time Auto makers have been chopping the time it takes to develop a new vehicle from 60 to 36 months to 24 — and today 16 months or less is being targeted.

Automakers are increasingly adopting modular system of vehicle architecture that is moving the industry to new levels. From an approach of iterative hardware of building a prototype and trying out the physical characteristics; math models are being used for characterization to predict hardware behavior. Besides characterization, mathematical models are being reused and the learning cycles from them are being adapted on other products. The combination of re-usable math models, 3D printing and rise of virtual reality tools have radically reduced time to prototype and product-development time for Automakers.

Another significant development has been in the process. From a step wise goal directed process involving a series of information acquisition activities and evaluation points, there has been a transformation in the approach of new product development process that has led to a dramatic push to simultaneous engineering; where designers, engineers, and manufacturing team all start working together from initial design for quicker rolling off the product line.

Typically, a car that is made out of 200 major components with 200 major processes to put it together; that's 40,000 things that Automakers need to perfect from scratch with FMEAs (Failure Mode & Effects Analysis), DVs (design verification), PVs (production verification), PFMEAs (Process Failure Mode & Effects Analysis), The efforts required to do all these concurrently in a conventional approach is a daunting task.

From a conventional approach, where a new concept would typically be tested through the production of several physical prototypes, Automakers are adopting VR technologies that provide a virtual environment for concurrent product design and manufacturing; thereby minimizing the time, effort and cost involved in creating and testing physical prototypes.

With an integrated approach of feasibility analysis, iterative design, systematic evaluation in product development to designing the most efficient assembly process, simultaneous Engineering has become close to reality.

By integrating VR tools, IoT, analytics, wearables and 3D printing, factories are gearing up to compress product development and production cycles and become responsive to external events as well as to changes in the supply chain and consumer demands.

As producers strive to reduce product development times to almost unheard of speed for quicker rollout of products; consumers’ choice is progressively shifing from standardized, mass-produced products, in favour of custom-made solutions that fit their specific needs. This consumer trend has left the producers faced with the challenge to expand their product lines to better address consumer preferences, or develop “mass customization” techniques. Businesses like in Automotive, demand for personalization runs counter to the dominat model of providing high volume products or services through mass distribution. Are the producers now ready to move from mass production to mass personalization?

Going by the rising trend of consumer choices, many of the high-end Car makers are now in the phase of incorporating VR technology into the showroom environment and allow customers to personalize their vehicle specs right in the dealership: everything from colors, to electronic systems, inlays, and interior leather.

With the current pace of technology development and greater opportunities to interact with consumers, striking a balance between Market pull and Technology push seems to be a certainty.

Can factories shorten production runs to “batches of one”, with high-quality, single-piece production at the economies of current mass-produced goods? Are economies dependent solely on labour arbitrage in manufacturing sector under threat? You may want to read the post, The Factory of the Future…Economies of Scale to Scope Economies

Will new business models shift from products to services? You may want to read the post, Creating business model...where digital meets physical