Current trends on ICT technologies for enterprise information systems

Abstract

As business conditions change rapidly, the need for integrating business and technical systems calls for novel ICT frameworks and solutions to remain concurrent in highly competitive markets. A number of problems and issues arise in this regard. In this paper, four big challenges of enterprise information systems (EIS) are defined and discussed: (1) data value chain management; (2) context awareness; (3) usability, interaction and visualization; and (4) human learning and continuous education. Major contributions and research orientations of ICT technologies are elaborated based on selected key issues and lessons learned. First, the semantic mediator is proposed as a key enabler for dealing with semantic interoperability. Second, the context-aware infrastructures are proposed as a main solution for making efficient use of EIS to offer a high level of customization of delivered services and data. Third, the product avatar is proposed as a contribution to an evolutionary social, collaborative and product-centric and interaction metaphor with EIS. Fourth, human learning solutions are considered to develop individual competences in order to cope with new technological advances. The paper ends with a discussion on the impact of the proposed solutions on the economic and social landscape and proposes a set of recommendations as a perspective towards next generation of information systems.

Introduction

Given the context of manufacturing companies, today’s consumer demands products of the highest quality accompanied by information and services which together constitute a holistic product experience. There is also a noticeable trend towards the consumer placing more value on the sustainability, pedigree and authenticity of products, making transparency along the stations of individual products’ lifecycles a growing concern for industry. Companies are increasing the number of new product introductions in the market leading to decrease the time-to-market and consequently to shorten the life cycle of the product itself. Moreover, sectors of manufacturing which have previously focused solely on the improvement of their products’ quality to remain competitive in the marketplace are turning towards emphasizing the after-sales market of their products to remain competitive. Especially the manufacturers of complex and high-value products are investigating new concepts of servitization and through-life engineering services based on the actual usage information of individual products by their customers. Services offered on that basis include traditional activities such as maintenance, upgrades, storage and refurbishing but also include ones provided in the virtual world integrated with social network services [74].

In order to meet these challenges, manufacturers need to take concepts such as item-level and closed-loop Product Lifecycle Management (PLM) [86] into consideration. This relies on the holistic availability of product-related data to all stakeholders throughout the entire lifecycle with the closing of information loops between the individual phases of the product lifecycle and between different IT layers, from data acquisition, through middleware and knowledge transformation to the business application layer. In order to consistently deliver the product experience demanded by the consumer, relevant information generated throughout the product lifecycle needs to be captured, managed and processed, for which different technological solutions have been proposed [52], [68], [86]. All of these have in common the augmentation of physical products with “intelligence” to facilitate data generation, processing and networking with other products, users and stakeholders throughout the lifecycle. That intelligence is implemented by different means, such as RFID, PEIDs (Product Embedded Information Devices), embedded systems, smart sensor systems, Single Board Computers (SBC), amongst others.

Increasingly, personal mobile devices (smartphones etc.) are capable of interacting with products and also generating and communicating valuable item-level product data in the context of the individual user’s product usage. These devices and the services running on them via apps are thus not only becoming increasingly valuable data sources, but also providers of context information. Most product-relevant data collected via personal mobile devices can and often already is directly connected to a number of Web 2.0 social network services. In Europe, 42% of citizens use online social network services at least once a week [46]. In a recent study of German social network services users and their time spent online, social networks were far ahead with Facebook leading with 56% of internet users being active there [10]. In other countries, like the USA the ratio is even more extreme [140], [142], [169]. Social network services are not only an accepted part of the daily life of most European citizens, but will also play an increasing role when it comes to future data generation (users) and integration (service providers). Access to the large amounts of product-relevant data continuously generated by users of social network services will only become more important over the coming years. It is crucial to understand the specific terms and conditions of the targeted social network service before starting to implement a solution of both, data capturing and integration.

Besides the emerging, dynamic and item-level data sources relevant to a product’s lifecycle, as described in the previous paragraphs, already existing conventional enterprise systems, databases and other data sources will retain their importance in the future. The landscape of relevant IT connected to the lifecycle of an individual product is generally distributed, heterogeneous, and, depending on the nature of the product itself, can be very complex. It generally involves many different enterprise systems distributed across multiple stakeholders, many of which either operate proprietary or legacy systems or can be small enterprises with no notable ICT infrastructure. Furthermore, stakeholders may unpredictably participate to the value chain, making the flexible addition or removal of data sources necessary where the data generated by these providers can be valuable to other stakeholders (e.g. in product design and test processes).

On the other hand, larger corporations will still demand “standardized” software and ICT packages that can efficiently be implemented in multiple sites and countries. The basic functionalities of these EIS need to be the same. Ideally, however, their context awareness is developed to the degree that the same software can be used on different sets of production machinery, for the production of different products and components and be integrated into the practices of different workplace cultures in different countries. These scenarios in combination make the importance of realizing interoperable and scalable EIS even more evident.

From this perspective, several roadmaps and surveys have been produced by various entities between 2004 and 2010, aiming at giving a prospective view on developments over the next 10–20 years. We take a look back at these predictions, to enlist what is already a reality and investigate how what has been realized impacts on or confirms what is still to be done.

For exploring the future trends of production systems, we have mainly considered the following roadmaps produced within 2000–2010:

• The main prospective roadmap of the European technology platform Manufuture: «Manufuture, a vision for 2020» [105], released in November 2004, and completed by a «Strategic Research Agenda» in 2006 [106]. These visions are relatively old, but have had a tremendous impact on the following studies.

• “Preparing for our future: developing a common strategy for key enabling technologies in the EU,” report of the European Commission [47].

• The roadmap of the IPROMS Network of Excellence [83], gathering more than 30 European partners from Research and Industry.

• The IMS (Intelligent Manufacturing Systems) roadmap “IMS 2020” [80], IMS being a well-known industry-led, international business innovation and research/development program established to develop the next generation of manufacturing and processing technologies.

• The roadmap of the European Commission “Factories of the Future” [48], including a “strategic sub-domain” on ICT-enabled intelligent manufacturing.

In addition we have also considered the survey [159], based on the following (and sometimes less well known) documents:

• For Europe, [55], [56]107] (as well as [105], already mentioned).

• For USA, the roadmap “Integrated Manufacturing Technology Roadmapping” [81] and a report on Manufacturing in US by the Dept. of Commerce US Dept. of Commerce, 2004.

• For Japan, a Delphi study on the Technologies of the Future [122], as well as macro-economic studies from [174], PricewaterhouseCoopers [139] and the World Bank [164] on the emerging technologies in 2050.

Two main competitive contexts emerge from the analyzed scenarios and roadmaps:

• A worldwide competition based on the design and management of very efficient global supply networks, in a context of increased uncertainty and instability (also linked to the political situation in emerging countries and to climate changes),

• The parallel emergence of local supply chains (at the regional, national or continental levels) in order to answer to political, ethical, environmental or supply reliability constraints.

These two opposite tendencies should coexist according to the type of product (raw materials and mass production in the first case; high-tech customized products and products reaching their end of life in the second case).

Facing the increased competition from developing countries, innovation is universally considered as a key point for sustainable competitiveness. Even if the conditions for innovation can hardly be formalized, its link with research, knowledge management, education and free exchanges is often underlined.

The industrial fabric being mainly composed of SMEs (Small and Medium Enterprises) all around the world, being able to disseminate new technologies within small companies, and being able to integrate them in global but efficient networks is considered as a major challenge.

The necessity to have a holistic approach on the life cycle of the products and organizations, taking into account the three dimensions of sustainable development (economical, societal, environmental) is also a common point in most of the studies. The societal dimensions of manufacturing (aging of workers and customers, job insecurity, teleworking), as well as environmental considerations (eco-design, economy of resources) may re-orientate classical themes on original topics.

The perception of increased product and services customization as a competitive advantage is universally shared. This induces new requirements for information and knowledge management for the development and production of more complex products. The product will be more active, during its manufacturing phase but also all along its lifecycle, thanks to ambient intelligence and connected devices technologies. This opens new perspectives, under the condition of being able to federate very different communication protocols. In order to make full use of ambient intelligence in a context of product customization taking into account environmental and societal constraints a deep re-design of production systems is required, including:

• more intelligent machines thanks to intelligent sensors and actuators, connected under the multi-agent or holonic paradigm, easier to operate and maintain, easier to reconfigure and in a better interaction with the operators through augmented interfaces

• more flexible workshops that can be re-organized in an opportunistic way

• organizations able to create at the same time stable partnerships on some high tech products, and ephemeral but efficient collaborations on short life products, exchanging knowledge (and not only information), using interoperable processes and information systems, benefiting intensively from external services accessible in SaaS (Software as a Service) mode, but also from distant human competences available as services.

The increased complexity of the products and organizations of the future, resulting from their required flexibility and resilience, requires new approaches for modeling “systems of systems” and corresponding risk assessment. The principles of the digital factory, allowing multi-scale simulation, should allow predicting the performance of both products and production systems. The coexistence of different actors (individuals and organizations) in all phases of the product and supply chain lifecycles creates a critical need in methods and tools for collaborative work and distributed decision making. A better interaction between partners within industrial processes should be made possible by the emergence of communities (e.g. partners, customers, workers), for instance using Web 2.0 tools, with enhanced exploitation of collective intelligence.

In relation with these topics, the explosion of use of ICT should allow to better perform classical tasks, especially in a new distributed context, but should also allow to completely re-think the interactions between actors, between the actors and the products and between the actors and the information systems of their organizations.

The recent McKinsey report on IT-enabled business trends for the next decade [175] illustrates how the roadmaps analyzed in the previous section have influenced the perceived future role of ICT in organizations. The report suggests ten trends in which ICT will enable companies to reach a new competitiveness:

• The “social matrix,” meaning that socially enabled applications will become ubiquitous, allowing liking, commenting, and information sharing across a large array of activities, both at the personal and professional levels (see the emergence of communities in the manufacturing roadmaps).

• The “Internet of All Things,” seen as an extension of the previous “Internet of things,” taking into account the unexpected proliferation of connected devices. Before challenging the imagination of engineers through new applications, this context creates a tremendous need for interoperability, at the technical but also semantic level.

• “Big data, advanced analytics”: also noticed in other whitepapers like [82], this trend can be considered as an “ICT oriented” interpretation of the knowledge based factory denoted by manufacturing roadmaps. Indeed, the real challenge of “advanced analytics” in the context of “big data” is not to process more information, but to create value from information, i.e. to extract and structure re-usable knowledge from data. The link between big data and the learning enterprise is for instance emphasized in the whitepaper [173]. In a distributed context underlined by the “manufacturing oriented” roadmaps, the link between big data and interoperability is also highlighted [114].

• “Realizing anything as a service”: IT clearly evolves from products (pieces of software that the users should install in their companies) to services, eventually opportunistically accessed. The recent arrival of ERP (Enterprise Resource Planning) products available as services like SAP “Business By Design” is a clear illustration of this trend, but sets again the problems of both semantic and technical interoperability.

• “Automation of knowledge work”: IT is expected to enable the automation of knowledge-based activities, similar to how automation technologies have in the past achieved to automate physical activities. In a more modest way, we can bring this trend close to the “learning enterprise”, in synergy with point 3.

• “Integrated digital/physical experiences”: as denoted by the manufacturing roadmaps, the “digital factory” should allow unseen possibilities in simulation. The McKinsey report points out that the interaction between the digital world and the human user will require new “natural” interfaces based on visualization (augmented reality for instance) or on gesture/voice interfaces.

• “Me + free + ease”: through this cryptic theme, the McKinsey report refers to the necessity of highly personalized customer service, characterized by extreme ease of use and instantaneous results, requiring to take benefit of point 6.

• “The evolution of commerce”: this more usual trend refers to the B2C generalization, setting again challenges in advanced analytics, interoperability and emergence of communities.

• “The next three billion digital citizens”: this is the number of newcomers (mainly from developing countries) expected in the Internet world in the next decade, setting the problem of the intrusion of the related techniques and resulting behaviors in societies with quite strong and distant traditional cultures.

• “Transformation of government, health care, and education”: Internet and IT are changing the daily life of persons and enterprises quite quickly. For McKinsey, the diffusion of these changes in the areas of responsibilities of the governments is much slower.

As seen in the previous sections, research in the field of ICT, which can be easily linked to the requirements of the “factory of the future”, involves a number of problems and issues. On the more specific domain of enterprise information systems (EIS), we have chosen here to structure them in Four Grand Challenges that need to be tackled and are discussed next. They are summarized in Table 1.

• Data value chain management

The immense amount of data relevant to an organization from distributed, heterogeneous data sources, will need to be made accessible in an appropriate way [175]. As seen in previous sections, novel approaches to flexible, virtual and semantic interoperability need to be developed to tackle this problem. Once the data is made accessible, it needs to be processed and analyzed to make use of it in value-adding processes and services. The high data volume, velocity, variety and veracity (4 Vs of Big Data) require novel approaches to data analysis and mining. Foremost business customers might not understand the potential benefit of sharing their data with others, or feel the risks associated with sharing data outweigh the potential advantages. A significant challenge is thus to develop incentive systems which make clear the benefits of sharing. As already seen, many products can and do generate data about their usage which can be shared with stakeholders. However, in many cases, owners and users choose not to do so. Business customers are often concerned about exposing operational knowledge which could be used to their disadvantage. Furthermore, open data and social media are increasingly being perceived as valuable sources of product usage information in the design and co-creation of products [133] and the provision of product service systems. These data sources can be used by producers to gather more detailed information about the actual use of a product by individual users, and fed back into different lifecycle phases to inform decisions throughout the lifecycles of the current or future product iterations. A significant challenge is consequently the development of secure infrastructures for sharing data with the different stakeholders HP, 2014 whilst retaining privacy and data security. This challenge needs to be addressed taking social, technical and legal considerations and solutions into account [7]. Secure infrastructures for big and open data sharing will consequently need to involve moving data security controls closer to the data store and data itself, rather than placing them at the edge of the network [153] while increasingly including technical means to create policy-aware data transactions [162].

• Context awareness

Interoperability between information sources, as depicted in the previous section, is a first condition for meeting the challenges of data value chain management. A second condition is to give access to the right information that supports a work task, a business decision or a cooperation process, which is often very difficult. In certain situations not all information provided by an information system is important and relevant to the end user. Modern enterprise information systems provide huge amounts of information and in those large volumes very often the user cannot find appropriate and important information at the right time. Moreover, in complex business environments sometimes users are not aware of the current situation which negatively influences the decision making process. It is therefore very important to provide the appropriate information to a user in a specific situation. Moreover the user also has to understand why the information provided is important which means that he/she has to understand the current situation or to be aware of the context in which it happened in order to understand the real meaning of the information. Therefore it has become crucial for enterprise applications to be aware of the context they are being used in.

• Usability, interaction and visualization

Appropriate means of interaction with Next Generation EIS are a further major challenge. On the one hand, the ubiquitous availability and use of computing devices in society mean that expectations towards user interfaces are very different to the past [175]. On the other hand, the vast amount of data and information to be visualized and manipulated by EIS in the future means that new and intuitive ways of presenting and interacting with that data will be required. Solving user interaction problems requires dealing with context awareness.

• Human learning and continuous education

Human learning is the process of identifying and implementing professional competences triggered by new scientific and technological knowledge and implemented in an industrial context to address new professional needs. Engineers and workers will need new life-long learning schemes to assist them in keeping up with the pace of technological change which requires a continuous update of the learning content, learning processes and delivery schemes of manufacturing education. ICT research outcomes of educational institutions are typically presented to the scientific community and are not directly accessible to industry. Uni-directional learning flows, such as learning via training, is surely important but not sufficient to cover the full cycle of enterprise knowledge flows. An upgrade in the learning mechanisms is urgently needed, placing the human at the center of the knowledge flow management process and bridging conventional learning with experiential, social and data-driven learning. Such an upgrade could eventually lead to facilitating transitions between different types of knowledge and enable novel technology/knowledge transfer schemes to have a significant impact on the ICT related innovation performance.

The remainder of this paper is organized as follows. Section 2 provides an outline of the contributions of the authors to the four grand challenges discussed above. Section 3 covers relevant literature review on the subjects related to these challenges and discusses the general state of research. Section 4 discusses the overall contributions highlighting the impact of these solutions on the economic and social landscape and concludes these works.