The lights-sparse versus the lights-out factory — Part 2
The opportunities presented by the concept of a ‘lights-out’ factory — one in which the requirements for human activity are so minimal that the facility can operate in the dark — was discussed in Part 1 of this article, along with some of the enabling physical technologies available today to make it potentially possible. But there are many challenges still to be overcome, from the emerging need for product customisation and varying labour and skills issues across the world, to entrenched mindsets.
In the end, it will be the ongoing increase in digitalisation and advanced software that will make possible lights-out — or at least lights-sparse — manufacturing.
Competing interests
While modern technological developments may accelerate the implementation of lights-out manufacturing, other recent trends and perennial market conditions make lights-out implementation more challenging. This is especially true for brownfield factories, where design and implementation of lights-out operations are complicated by the necessary transition from existing manufacturing approaches. Greenfield factories, on the other hand, can leverage technologies right away to prepare the facility for the challenges listed below, building the facility with factory floor modularisation, automation, 5G, AI/ML, other enabling digital technologies and a data-driven mindset.
Design anywhere, build everywhere is the latest mantra associated with the trend towards manufacturing globalisation. By making deliberate choices about where and when to run production, manufacturers expect to reduce manufacturing costs and accelerate time to market. They want to make production decisions based on location, on-site material inventory, machine capacity, skilled labour availability, shipping requirements — any parameter that impacts timely, efficient and cost-effective production. For these global manufacturing enterprises, lights-out manufacturing adds a dimension of complexity to any design anywhere, build everywhere decision matrix.
If a lights-out or lights-sparse factory is one of the build everywhere options, for example, considerations regarding skilled labour may be replaced with those regarding the manufacturing equipment’s capabilities for autonomous operations, as well as the equipment’s or work cell’s ability to self-configure or reconfigure based on the product variant to be produced. Additionally, even if a dark factory, or lights-out operation within a conventional factory, is capable of carrying out a particular production run at a lower operational cost than a fully conventional factory, does that location have access to needed resources? Digital tools are essential not only for the planning and scheduling decisions to be made in such circumstances, but also for communication of design, supply, production engineering, quality requirements and inspection/test mechanisms and feedback of production information.
Mass customisation and individualised production are growing requirements across nearly every manufacturing sector, from specialty food and beverage items to new automobiles. Consumers want individualised products, but at the prices they would pay for mass-produced goods. To keep up with these demands, manufacturers must accommodate short innovation cycles with more complex products, and as a consequence, they need to more closely integrate modelling, engineering and simulation with manufacturing production. This means that the manufacturing floor must be more flexible than ever before — but lights-out manufacturing operations are more suited to long, steady production runs of the same product.
Ultimately, lights-out manufacturing and flexible manufacturing are not mutually exclusive; in fact, digital manufacturing tools available today are capable of addressing the need for flexibility and realising lights-out where it is feasible. Until a manufacturer has outfitted a factory with modular production equipment able to reconfigure itself, however, these digital tools will have to treat mass customisation and lights-out manufacturing as competing interests.
Trade-offs between labour and automation also directly confront manufacturers considering lights-out manufacturing. Generally, humans are flexible but more error-prone and slower, while automation is less flexible but faster and less error-prone. With the design anywhere, build everywhere and mass customisation trends described above, flexibility may be the factor for manufacturers deciding between lights-out and conventional operations. Again, digital systems for manufacturing engineering can facilitate such decision-making today with tools such as plant and process simulations.
Entrenched mindsets and human inertia may also keep some manufacturers from initiating moves towards lights-out manufacturing. According to the Gartner study mentioned in Part 1, advancements in technology will make lights-out production processes a reality in 2025. But the lights-out smart factories must balance between augmented and automated processes, organisational and technical challenges. For manufacturers with the foresight to explore and prepare for lights-out manufacturing, there is an opportunity for a powerful competitive advantage.
The path to lights-out manufacturing traverses every point of product life
How does a forward-thinking manufacturer prepare for lights-out manufacturing? What are the prerequisites? Once a decision to invest in lights-out or lights-sparse operations has been made, the key to implementation is how a company will leverage digital manufacturing tools to support crewless areas of the shop floor. Implementation of lights-out manufacturing, as discussed, can be undertaken incrementally, with certain operations or certain time periods of a production run targeted, one at a time, for crewless operation.
For example, in many complex manufacturing operations, inspection of work-in-progress is often performed manually by human inspectors. This may be true even when manufacturing value-added operations are highly automated, and this makes manual inspection a bottleneck against productivity gains. Alternatively, automated inspection is a type of lights-out or lights-sparse undertaking. The inspection system captures data or images of as-built features (or anomalies) and analyses it, referring to design data as the standard. When non-conformances occur, a lights-sparse operation might send an alert to a human operator for corrective action; a lights-out inspection system would include automated resolution of non-conformances. Beyond this detection/correction activity, automatic inspection data may be analysed to generate manufacturing intelligence that feeds continuous improvement efforts. When embedded with AI/ML-based algorithms, an unsupervised and supervised learning mechanism can be built to create possibilities for needed self-resolution/prescriptions.
Preparation for each lights-out manufacturing operation within a production facility must include the closing of gaps between virtual and real production, between the digital and physical factory. After all, without human intervention, production machines will take only those actions they are programmed or digitally directed to take. To realise lights-out manufacturing, then, manufacturers must consider how the virtual and real worlds will interact throughout a product lifecycle.
When actual operations are no longer attended by human operators, digital tools must provide manufacturing overseers with full, interactive visibility into each lights-out function. Visibility is achieved by orchestrating operations and collecting critical data to implement needed integration for lights-out operations. Simple products made on one or two machines require relatively uncomplicated communication from design to manufacture; but complex products involving multiple operations and machines require integration from orders to scheduling to production and delivery. Additionally, digital tools must enable dynamic responses either to changes in the supply chain or to upstream change management, along with seamless virtual validation for flawless lights-out production.
And then there is the question of efficiency: the ideal is for each machine to remain busy. To achieve this level of efficiency when one or several production steps will take place without human intervention, the digital manufacturing processes account for different time requirements of different operations, the need to minimise process changeovers, complex coordination of material movement and many other factors that are closely tied to the automation layer. By anticipating and addressing these issues digitally, lights-out operations replace human flexibility and spontaneous responsiveness with data-driven manufacturing. That is, digital systems provide real-time data insights that are fully correlated across processes, enabling automated predictive and prescriptive responses.
The digital twin: the workhorse of lights-out manufacturing
Central to the digital tools that facilitate lights-out and lights-sparse operations is the digital twin, which is the virtual representation of a physical product and associated processes, from conception through lifetime product performance in the hands of the end user.
Visibility of the digital twin is provided by a data infrastructure known as the digital thread and its continuity along the value chain. The digital twin provides the format into which design, engineering, manufacturing and field data are organised in coherent, accessible and useful models; and the digital thread (once fully built out) creates a collaborative, connected information conduit across product design, production engineering, manufacturing execution, automation and data-driven intelligence.
A comprehensive digital twin captures the full product life and is needed for fully automated production in a dark factory, but more specific types of digital twins are key to the incremental implementation of lights-out manufacturing operations. There are four key types of digital twins for manufacturing operations: the digital twins of product, production facility, production and performance. As a precise digital model, each digital twin displays development through the entire lifecycle and allows manufacturers to predict behaviour, optimise performance and implement insights from design and production experiences.
- The digital twin for product provides the virtual-physical connection that helps manufacturers analyse how a product performs under various conditions and make adjustments in the virtual world to ensure that the physical product will perform as planned.
- The digital twin for production facility helps manufacturing engineers and planners navigate through the facility in digital representations and visualisations.
- The digital twin for production helps manufacturers address the process engineering changes needed to accomplish lights-out operations.
- The digital twin for performance helps provide timely insights from lights-out processes that can be used in advanced analytics for data-driven orchestration of ongoing lights-out activities.
With the comprehensive digital twin, manufacturers gain foresight by simulating product, people, processes and resources in the virtual realm before implementing production on the manufacturing floor — a key first step in their journey towards lights-sparse processes. They gain additional insights by matching real-world to predicted performance, and they can use those insights to drive continuous improvements. While the digital twin is advantageous in any manufacturing environment, it is essential to lights-out manufacturing.
Digital tools are essential to lights-out operations
Digital enterprise software solutions are the means by which the digital twin and the digital thread are created, accessed and utilised to implement manufacturing operations of all kinds, including lights-out operations.
Within the digital enterprise, the central support of lights-out and lights-sparse manufacturing endeavours is provided by a digital manufacturing solution, including manufacturing engineering (with simulations), virtual commissioning and manufacturing operations management (MOM) software. Digital manufacturing software serves as the bridge from the virtual world of product ideas, computer-aided design (CAD) and planning to the real world of production. It is the basis for digital continuity between innovative product designs and best-in-class product and production performance. These digital tools ensure that quality and efficiency are built into the manufacturing process and are proactively and systematically enforced.
Just as manufacturers may approach lights-out manufacturing on an incremental basis, the same is true of their implementation of digital manufacturing. Manufacturers can employ digital manufacturing subsystems to implement the digital twin and the digital thread in a modular, stepwise fashion, with ongoing returns on investment that support the next module or step. From their current level of digitalisation, manufacturers can incrementally integrate legacy and custom functionality systems.
Digital solutions will usher in new capabilities in lights-out and other future factories
Today’s digital technology is capable of orchestrating lights-out manufacturing processes, work cells and production shifts. The move to greater lights-out implementation, and, ultimately, the widespread growth of full lights-out factories, seems likely to take place through a slow evolution of manufacturing approaches — but the technologies enabling this move are already on the horizon. A marriage of digital and physical manufacturing technologies will enable increasing flexibility, to a point in the future where lights-out factories are able to accommodate mass customisation and design anywhere, build everywhere manufacturing approaches.
Two key developments will open the door to this growth: modular automation and the ability to self-configure at a plant-wide level.
Conclusion
As digitalisation follows automation into the Fourth Industrial Revolution, concepts like lights-out manufacturing will be realised. These developments are not a mere fulfilment of some sci-fi vision of the future from the past century. Instead, they are the outcome of collaborative efforts among manufacturers, automation specialists and digitalisation software suppliers seeking to achieve the perennial goal of manufacturing: an increase of productivity, efficiency, speed and quality, resulting in higher competitiveness for companies on their way to the future of industry.
Anticipating maintenance problems with predictive analytics
By utilising predictive analytics, process manufacturers can predict failures, enhance...
Air-gapped networks give a false sense of security
So-called 'air-gapped' OT networks can still fall victim to cyber attacks, so what is the...
Maximising automation flexibility: the ISV-driven approach
Vendor lock-in has long been a significant barrier to innovation in the industrial sector, making...