A re-engineering perspective to assembly system development

Citation

Onori, M. (2005), "A re-engineering perspective to assembly system development", Industrial Robot, Vol. 32 No. 5. https://doi.org/10.1108/ir.2005.04932eaa.002

A re-engineering perspective to assembly system development

A re-engineering perspective to assembly system development

Keywords: Assembly planning, System development, Design for Assembly

The desire to conquer markets through advanced product design and trendy business strategies are still predominant approaches in industry today. In fact, product development has acquired an ever more central role in the strategic planning of companies, and it has stretched its influence to R&D funding levels as well. It is no surprise to see that many national R&D project frameworks within the EU today are dominated by product development topics, leaving production engineering, robotics, and systems sidelined. The reasons may be many but, unfortunately, product development and business strategies are not, and will never be, the only core competences to focus upon.

Design for Assembly techniques and methodologies have been in use since the early 1980s. Not only has it become a known methodology, it has led to the spawning of many other approaches. Nonetheless, their use has not truly been as widespread as initially expected. If DFA, DFA² and other methods have not acquired such a widespread usage within industry it is often because product designers state that methods limit their creativity. This is an absurd statement; since any respectable artist or architect will clearly point out that art is not created from pure blue-sky creativity. Creativity is a necessity, but without methods and guidelines, there would not be so much art out there. A creative situation without rules and methods is comparable to a mathematical situation with an unlimited number of combinations and permutations. Faced with such a dilemma, most people will require immense lapses of time to produce, if at all, a minimum number of results. The quality of these results, being placed on an unregulated panel, will be difficult to assess. In fact, methods and guidelines are a fundamental prerequisite for effective creativity. Without a framework to work within, it is difficult to achieve what is required. Product design methods are definitely required, but they must be directed at effectively interacting with the most fundamental problems: sustainable, cost-effective production.

In a more provocative tone, one may add that production systems which have truly changed the world, such as Ford's T- Model assembly line, Toyota's Lean Manufacturing, and Sony's Smart Cells, have all placed the production system at the center of attention, not the product: the product design, in fact, was carried out on the basis of the production system constraints. The issue is that these production systems were so cost-effective that they enabled the launch of highly novel products at reduced prices. A quick look at the success achieved by Surface Mounted Technology (SMT) providers gives us an important clue: if the assembly processes dictate some constraints upon the product design, standardisation and serious cost-reductions become a reality. Risks are minimised, product design finds suitable frameworks to work within, and the market literally explodes.

The major, and ever more predominant, argument against such ideas is that they neglect the fact that marketing is the strongest driver and that selling the product is the most important factor, not producing it. This is short-term thinking at one of its most extreme manifestations. Selling a product is the most important factor, of course, but its price, performance and availability are tightly linked to how cost- effectively you may produce it. Owing to such marketing and product-sales oriented approaches, product design has achieved an all too predominant role in the way we consider company strategies. In fact, product design is often the domain which assumes the leading role in the production life cycle planning. Even the assembly system developers, commonly focussed on the technological issues alone, are increasingly focussing on product design aspects and business strategies. Unfortunately, by letting final production aspects assume a secondary, final role, this continues to lead to assembly systems that are too dedicated and short-lived. Hence, by being inevitably caught up with very high production costs and risks, outsourcing of assembly becomes their predominant solution: cheap, low-wage manual labour offered by Contract Manufacturers. This may turn out to be a fatal mistake, and underscores the short- term perspective of such strategies: when outsourcing production, you lose control of the processes which may render your product unique. In fact, by outsourcing assembly, we are losing the main source of information to a possible reversal of this situation: process knowledge, often best held by the operators at shop-floor level, is being exported to low- wage nations. Furthermore, by handing the problem over to highly flexible machines such as humans solves the immediate problems only temporarily. They adapt, learn and solve the situation, but the knowledge acquired, experience and detailed process know-how are never recorded or exploited in future cases. If we add outsourcing to the equation, we clearly see what type of knowledge we are giving away for free.

A potential reversal of such a dilemma may be in changing the strategic perspective from system installation (known product launch) to re-engineering (future product launch). Figure 1 tries to clarify this.

Figure 1 Focussing on re-engineering can extend the development horizon and highlight production problems

The ideal scenario is to be able to do this in an existing assembly system. As of today, this often requires the extensive involvement of the system supplier, and vast amounts of re- programming and equipment modifications. However, if the industrial and academic communities would start to focus on the re-engineering phase (shaded area), the development horizon would be extended by many years, and the real production problem area highlighted. The final objective being effective (cost and performance), re-configurable production system designs. If this is achieved, the product designs can always be accounted for in production by creating a strong link between system design and product development: one basically knows which processes can be accounted for, and how. Assembly or production system constraints will be placed upon the product design!

In terms of applying a re-engineering perspective, what is required is not a solution which tries to accomplish all of the envisaged assembly needs within a closed unit (Flexible Assembly systems) but, rather, a solution which, being based on several reconfigurable, task-specific, process-oriented elements (system modules), allows for a continuous evolution of the assembly system. This new approach is known as Evolvable Assembly Systems (EAS), which is based on the modularity of assembly components which can subsequently be used interchangeably to create specific assembly systems whenever a business opportunity arises: in effect, there are no defined systems within EAS, only process- oriented modules.

This, however, brings us to a discussion of what a system is. In fact, a system's capability is not the sum of the capabilities of its parts. It could be a completely different scenario altogether. The issue being raised here is that EAS will consist of a vast range of inter-connectable modules, including robots. When a system will be created according to the EAS principles, the resulting capability of the sum of the modules will not be so easily predicted. What happens, in fact, when many small entities are brought together, is that new, unthought-of capabilities may emerge from this coalition. This calls upon the principles of emergence, since the capabilities being brought together may also be viewed as particular functionalities (skills) being offered by each module. A very important point to highlight is the fact that the properties that emerge out of the interaction of the modules that compose the assembly system represent just the complex functionalities (complex skills) of the system we want to create. The interesting conclusion is that we can generate any functionality we want as long as we provide the control architecture able to accommodate for the interaction of modules. From a robotics point of view, the industrial robotics community needs to adapt to this, and probably start paying more attention to what the “mobile” robotics R&D is achieving. The day when an IRb can freely communicate with its grippers and feeders may be far away, but it would certainly lead to emergent behaviour and a drastic reduction in re-programming.

The arguments brought forward have attempted to highlight the importance of the role of production. In a world in which quick success is more important than long- term gains, these words may fall on deaf ears, but the fact remains that the most difficult work remains to be done if full control of our assembly processes is to be achieved: structure, formalise and control the assembly processes, which range from arc welding of ship hulls to cryogenic gripping of microstructures.

M. Onori Associate Professor at the Department of Production Engineering, The Royal Institute of Technology, Stockholm, Sweden