Elsevier

Computers in Industry

Volume 60, Issue 3, April 2009, Pages 217-228
Computers in Industry

Intelligent products: Agere versus Essere,☆☆

https://doi.org/10.1016/j.compind.2008.12.008Get rights and content

Abstract

The notion of an intelligent product places suitability for integration firmly on the research agenda. Indeed, this paradigm aims for coordination and integration on a scale that is unseen until today. Note that this applies particularly where it concerns the core businesses of companies. For that reason, this paper presents a novel concept, the intelligent being, as a vehicle to achieve suitability for integration. The concept is applied to intelligent product instances, intelligent product types and intelligent resources alike. The paper identifies: (1) which services and functionalities can be offered by intelligent beings, (2) what are suitable candidate intelligent beings, and what are not. The paper shows that intelligent beings can reflect a corresponding reality – a product – in its current, past and future states. It thus argues for a role of the intelligent being that is analogous to what maps contribute in navigation systems (and may become as important).

Introduction

Technological developments, commonly associated with RFID1, have created the opportunity and ambition to organize manufacturing and logistic activities in radically innovative manners. Note that this ambition implies coordination and integration on a scale that has never been attempted before. In this setting, this paper addresses the integration issue related to the notion of an intelligent product.

Suitability for integration, in this paper, concerns:

  • the use of existing components and subsystems as well as;

  • the use of independently developed components and subsystems;

  • to build larger systems (as many as possible);

  • while respecting the integrity of these components and subsystems.

In information technology, “to respect the integrity of components and subsystems” implies rigorous restrictions. Validation represents a major part of any software's value. Changing a single line of code – in manners not anticipated by its designers – invalidates a component and therefore breaks this component's integrity. Software integration has to occur additively; it is restricted to adding wrappers, interconnections, etc. When a development requires a redesign of the components themselves, it is no longer integration but a cannibalizing activity that results in new components. Indeed, the resulting components cannot replace the original components in pre-existing installations. In summary, components and subsystems that are suitable for integration can be employed without modifications to build larger systems. And, this has to be true for a wide range of possible larger systems.

Seen from a different perspective, components and subsystems need a design that avoids the introduction of constraints that may hamper future integration. Indeed, a design process can be seen as the stepwise introduction of constraints through design choices [1]. The inapt introduction of constraints reduces the suitability for integration. This is especially true for design choices that are arbitrary (e.g. on which side of the road cars must drive), and not for choices that are well-founded (e.g. trucks and cars must drive on the same side). A properly designed intelligent being only requires well-founded choices.

Moreover, integration problems arise at – qualitatively – different levels within manmade systems and components. For instance, in electrical devices, the interface to the grid has three such levels:

  • the mechanical interface or plug;

  • the voltage;

  • the frequency and/or shape of the electrical signal.

Connecting or integrating incompatible systems becomes a lot harder while descending this list. Many households possess an instance to handle the first case: an adapter plug. The second one is already bulky and has a non-negligible price tag: a voltage transformer. The last two can be handled within microelectronics but at kilowatt levels the solutions probably cost more than the device itself.

Likewise, connecting mechanical drive shafts goes from trivially simple via complex to nearly impossible. Incompatible rotational speed ranges and torque limits, but compatible power ceilings (i.e. torque limit multiplied by speed limit), necessitates a gearbox. Power incompatibilities reduce the overall system to the lowest power across the components and introduce a risk of damaging the equipment. As an even worse case, mechanical oscillations and instabilities may invalidate the internal component designs.

These two examples serve to infer that software integration efforts commonly address the first and easiest level: making software interfaces compatible avoids the need for an “adapter plug.” It is useful but of comparatively minor significance. More challenging for software engineers and system developers is the suitability for integration at the more serious levels. Note that this involves application domain expertise. It becomes a multidisciplinary undertaking.

This paper addresses suitability for integration at these more serious levels. It does not present a panacea. Instead, it identifies which components and subsystems can be designed such that they are highly suited for integration. And, the paper discusses how to design such components. Furthermore, the paper proposes to develop such components, suited for integration, first and to introduce the remainder of the system second. The motivation to present this research in an intelligent product context resides in its aptness as a foundation for components that are suited for integration. Intelligent products and intelligent beings are a natural match.

This paper first motivates the use of sophisticated software technology – active objects – to implement intelligent products. Next, it introduces a new software concept – the intelligent being – that delivers the suitability for integration. Third, the paper addresses which components can be developed in this manner (and which functionalities cannot). Fourth, the discussion reveals that the services offered by such intelligent beings include short-term forecasting, which appears at first to be out-of-reach for these intelligent beings. This chapter makes a case that the intelligent beings, in the setting of intelligent products, are able to contribute to an extent that maps contribute to navigation. Finally, conclusions are given.

Section snippets

Intelligent products are active objects

Excessive simplifications in representations, models and theories are a source of never-ending tuning, imprecision, complicated stopgap procedures and degraded performance. Consider building a map of Europe on which distances and surfaces must be correctly indicated. A more expressive modeling technique, supporting the representation of curved surfaces in three dimensions, renders this a straightforward task. Forcing such model onto a flat surface complicates matters significantly: projections

Suitability for integration: intelligent beings

This section identifies what kind of intelligence may be added to a product without losing the ability to effortlessly integrate the resulting intelligent product into any overall solution or system. Note that this severe integration requirement will restrict the kind of functionality that may become part of this intelligence.

Which Intelligent Beings? Critical mass versus complexity

The previous sections argued that the intelligent products need to be active intelligent beings first accompanied by intelligent agents next. The intelligent agents handle the decision-making functionality for which the corresponding reality fails to provide shelter. The intelligent beings provide the ease-of-integration and minimal maintenance because they are sheltered by a corresponding reality, which is coherent and consistent. Nonetheless, this leaves an open question: which parts of

Intelligent resources, intelligent products and short-term forecasting

Intelligent products are a combination of an intelligent agent and an intelligent being. The agents are exposed to the dynamics of the environment and their life cycle (longevity) and applicability fails to coincide with a corresponding part of reality. Their specialty is decision-making and achieving objectives. In contrast, the intelligent beings delegate decisions to their associated agent and are thus protected by a corresponding reality. The latter makes an intelligent being the preferred

Intelligent products

The diverse research results concerning intelligent products address various concerns and viewpoints. For instance, McFarlane et al. [18] address manufacturing control, providing a demonstration how Auto-ID systems facilitate the development of a Holonic assembly cell. This basically reveals the potential of the intelligent product concept without providing reusable components in a systematic manner. Wong et al. [19] address supply chain issues, and provide a systematic list of services and

Conclusions

Recent technological developments, commonly associated with RFID, have created the opportunity and ambition to organize manufacturing activities in radically innovative manners. Here, manufacturing must be interpreted in a very broad sense, i.e. including logistics, after sales services, product end-of-life management, etc. Importantly, this ambition implies coordination and integration on a scale that has never been attempted before: design, manufacturing, logistics, maintenance and supply

Acknowledgement

This paper presents work funded by the Research Fund of the K.U.Leuven – Concerted Research Action on Autonomic Computing for Distributed Production Systems.

Paul Valckenaers received the applied mathematics engineering degree, the computer science engineering degree, and mechanical engineering Ph.D. degree from the K.U. Leuven, Belgium. He is with the Mechanical Engineering Department of the K.U.Leuven. Since 2007, Paul Valckenaers is a Fellow of the K.U.Leuven Industrial Research Fund, and focuses on valorization-oriented R&D.

References (26)

  • H. Van Dyke Parunak, D. Weyns (guest editors), Special Issue on Environments for Multi-agent Systems, Autonomous Agents...
  • S. Bussmann, An Agent-Oriented Design Methodology for Production Control. Ph.D. Thesis, The University of Southampton,...
  • H.A. Simon

    The Sciences of the Artificial

    (1990)
  • Cited by (42)

    • Towards the deployment of customer orientation: A case study in third-party logistics

      2019, Computers in Industry
      Citation Excerpt :

      In an industrial context, an intelligent product refers to a physical order or product instance that is linked to information and rules governing the way it is intended to be made, stored or transported that enable the product to support or influence these operations [9,13,8]. For their implementation, intelligent products require digital elements (for information storage and decision support)—such as software agents [14,15]—to be connected to physical products (or orders), via means such as RFID, bluetooth, WiFi, QR, automated ID systems [16–20]. The paradigm is aligned with cyber-physical systems and aims to improve the responsiveness, flexibility and reconfigurability of industrial systems [21–24].

    View all citing articles on Scopus

    Paul Valckenaers received the applied mathematics engineering degree, the computer science engineering degree, and mechanical engineering Ph.D. degree from the K.U. Leuven, Belgium. He is with the Mechanical Engineering Department of the K.U.Leuven. Since 2007, Paul Valckenaers is a Fellow of the K.U.Leuven Industrial Research Fund, and focuses on valorization-oriented R&D.

    Bart Saint Germain received the software engineering degree from the K.U. Leuven, Belgium. He is a candidate for the Ph.D. degree in mechanical engineering from the K.U. Leuven. Since 2002, he is with the Mechanical Engineering Department, division PMA, of the K.U. Leuven.

    Paul Verstraete received the software engineering degree from the K.U. Leuven, Belgium. He also obtained a degree in industrial economics at the same university. He is a candidate for the Ph.D. degree in mechanical engineering from the K.U.Leuven. Since 2003, he is with the Mechanical Engineering Department, division PMA, of the K.U.Leuven.

    Jan Van Belle received the electro-technical engineering degree from the K.U. Leuven, Belgium. Since 2006, he is with the Mechanical Engineering Department, division PMA of the K.U. Leuven.

    Hadeli received the industrial engineering degree from the Parahyangan Catholic University, Indonesia. He obtained the master degree in industrial engineering and engineering management from the Institute Technology of Bandung, Indonesia and master degree and Ph.D. in engineering from K.U. Leuven, Belgium. Since 2007, he is a Research Scientist with ABB Corporate Research.

    Hendrik Van Brussel is full professor at the Faculty of Engineering of the K.U. Leuven. He received his M.Sc. EE and Ph.D. degrees from K.U.Leuven, Belgium. He is fellow of SME and IEEE and he received honorary doctor degrees from RWTH, Aachen and from the ‘Politehnica’ in Bucharest. He is a Member of the Royal Flemish Academy of Belgium for Sciences and Fine Arts, past President of CIRP and Foreign Member of the Royal Swedish Academy of Engineering Sciences.

    Agere: Latin for “to act,” still discernable in the word “agent.”

    ☆☆

    Esse: Latin for “to be,” still discernable in the words “essence” and “essential.”

    View full text