UNION INTERNATIONALE DES TELECOMMUNICATIONS INTERNATIONAL TELECOMMUNICATION UNION UNIÓN INTERNACIONAL DE TELECOMUNICACIONES Remarks to Nordic Government Computer Centres EDIT Conference Friday, March 5, 1993 International Standards and EDI Pekka Tarjanne ITU Secretary-General The electronic exchange of information - whether messages, textual documents, well-formatted trade documents, or other kinds of files, is increasing daily in volume and in the number of organizations taking advantage of this technology. These electronic methods result in increased speed of information flow and lower human processing costs; both are of importance in today's competitive world. For governmental institutions, the incentive is better service with an increased volume of work within strict budgetary constraints. Just as commerce itself is internationalized, electronic data interchange (EDI) must cross national and regional boundaries. One benefit of using EDI is near instantaneous communication. Of course, this is not unique to EDI - telex and even the telegraph allowed rapid communication among distant trading partners. But EDI lends itself to automated procedures that speed up the actions to be taken upon receipt of a communication. When EDI uses a store-and-forward mechanism, such as X.400, processing can go on asynchronously without concern for time zones or the immediate availability of equipment or personnel at the partner's site halfway around the world. Within the European Community, EDI has been the focus of particular attention - and funding - by the Commission of the European Communities. The Community's Trade Electronic Data Interchange Systems (TEDIS) initiative, originally an ECU 5.3 million two-year program in 1988-1989, was extended to a second phase (1991-1994) with a budget of ECU 25 million. The Community's stated objective is to "create an EDI culture in the European Single Market." Early EDI applications usually involved closed-user groups, for example: a major automotive manufacturer and its principal suppliers; the banks who set up SWIFT for electronic funds transfers (EFT); or SITA which links together the world's airlines. As EDI comes to be more widely used, instead of restricting EDI to closed, sectorial applications, generalized EDI implementations make sense and can be based upon commercially available products and services. When setting up EDI applications today it is wise to think in the broadest international terms. Even where present trading patterns may show a preponderance of regional partners, the availability of a standardized worldwide communications network reduces the marginal cost for adding distant partners, while laying a sound basis for expansion as trading patterns themselves evolve following political changes and the evolution of multi-lateral trade arrangements such as the GATT. X.400 electronic mail is reported to be available even in some countries of the CIS. It is also strongly preferable to select standards which are widely applied, rather than fashioning an ad hoc implementation to address specific sectorial requirements today. Adopting international standards does not preclude customization, since the most widely applied standards include provisions for designing messages needed for particular applications. The worldwide telecommunication network and ever more unified worldwide market for telecom services are based on standards. Standards cover a range of subjects and components e.g., services per se, quality of service, as well as the interfaces, telecom equipment, signalling systems, accounting arrangements....etc. I am sure that as Information Systems people, I don't have to convince you of the value of standards today. A value that increasingly is of benefit to users as well as suppliers in a worldwide market. STANDARDIZATION PROCESSES Standards bodies today form a complex, non-hierarchical matrix of units where the information is constantly being transferred, compiled and adapted by numerous different organizations. With relatively few exceptions, manufacturers and service providers need standards within time frames that are today measured in months rather than years. Even then, it is often necessary to adjust specifications to align with constantly advancing capabilities in basic technology implementations in processors, memory, and transmission speeds. Each additional month represents major losses in opportunities and major costs. The ITU, as a specialized agency of the United Nations, represents the interests of a broad constituency - including many countries that are unable to gain effective access to information or to participate in the meetings of standards bodies. We feel an obligation to be an advocate of any measures that reduce those barriers. No one standards organization can do all the work in the required time frame with the required specification and with the necessary service to local constituents. As a result, there is a layering effect where global bodies (like ITU and ISO) make general - often abstract - standards or models with many options. Further amplification and internetworking among standards bodies is increasingly crucial. Tasks may be appropriately divided by subject, as is the case in EDI matters, where the United Nations Economic Commission for Europe (ECE) Working Party 4 is responsible for the content-related EDI message definitions. There is close cooperation between ECE and the ISO/IEC JTC1 for protocols and syntax, and communication matters are referred to CCITT (from March 1, TSB). The relationships among the standards groups can be fairly complex, as illustrated in the table "Current Taxonomy Matrix" (below) which shows the division and sharing of responsibilities among the standards groups involved in EDI. It is important to provide a platform for global cooperation and close collaboration within the standard making process. If you look at Europe, North America and the Far East, there are active and efficient regional bodies. We have been used to national standards for a long time, and now there are regional and global standards as well. In this structure, each level should support the other with a view to reaching an optimal solution. With the internationalization of the telecommunications market, it's in everybody's interest to have global standards wherever they are needed. Frequently a harmonization process takes place in which proponents of regional standards may come together in international bodies, like TSB, to create an international standard melded from elements of the regional standards; for example, the incorporation of American SONET elements in CCITT's Synchronous Digital Hierarchy (SDH). Sometimes a regional standard, like the ANSI X12 EDI standard will have served its purpose for a time, and a consensus evolves that the internationally sanctioned standard provides a superior basis for the future. I understand that a recent poll has shown 75% agreement with phasing out of X12 in favor of EDIFACT by 1997 in North America. Another problem being faced today is that often standards may not have been tested to see if they actually work even as they are adopted. It is usually left to regional and/or national bodies and/or individual providers to subsequently develop, flesh out, and test more detailed, implementable standards, or implementation profiles which specify a subset of the options given in the more general international standard. Separate conformance testing standards bodies have sprung into existence over the past few years to focus on the testing requirements. In addition, as we have found in the recent problems with Signalling System 7, some means of sharing test information and operational experiences may be increasingly necessary as complex protocols are introduced in advanced networks. Existing or emerging new regional and global policy requirements based on antitrust and trade principles are reflected in legally binding agreements and laws. All of these developments require that the standards making processes be transparent and open, provide prior notice, and easy access to the resulting standards and drafts. The ITU, based on the recommendations of its High Level Committee (HLC) to review the structure and functioning of the ITU, agreed on action inter alia for restructuring and speeding up its standardization activities. Like the operators, we should become more user-oriented. Partly because it has sometimes been in its interest, the ITU is little known in the private sector. As a result, people have seen the ITU as a bureaucratic intergovernmental organization which was building barriers to free trade. The HLC report reflects the changing telecommunications world and identifies the improvements needed in the structure and workings of the ITU. The ITU has never been solely intergovernmental, and private companies have always had a part to play. Today, in its standardization activity, more than 85% of the input comes directly from private companies - which participate in the work of the TSB in their own right. We have recently set up a World Telecommunication Advisory Council (WTAC) in line with the HLC Report, where the CEOs and senior management of leading telecom companies are represented and can give good advice to the ITU and a Telecommunication Standardization Advisory Group (TSAG) is to be established as a new instrument to improve the efficiency of global standardization in the future. We also need close collaboration with the "end users" of telecommunications. The user may be someone in a remote rural region that simply wants the lowest cost telephone service available or it may be a large multinational corporation which wants to interconnect its operations through an integrated broadband private network which it will maintain. Large business customers are typically highly knowledgeable users, often possessing their own advanced facilities. They generally seek the ability to access and manage public network resources to meet their specialized needs. These customers often have the ability to adjust their operations to take advantage of lower telecommunication costs. On the other hand, there is the average individual who also wants nothing more than convenient low-cost telephony service or some easy, minimal information system access. Even here, there may be diversity between the urban customer and one in some remote region - differences that invoke significantly different network strategies. Speaking of user participation and the cooperation of participants with different interests, there is the example of the CCITT Associate Rapporteur Group for EDI and X.400 which developed the CCITT X.435 and F.435 recommencations dealing with EDI messaging. It was composed of a combination of X.400 experts, EDI experts and EDI practitioners. The X.400 experts tended to view the world as concentric circles with EDI completely contained within X.400. That is, EDI was considered to be one type of data to be transmitted within X.400. This view tended towards imposing X.400 conventions - like the use of Abstract Syntax Notation 1 (ASN.1) to encode EDI information. EDI practioners tended to view the world as illustrated below: That is, X.400 was considered to be one type of transport mechanism for EDI data. This view tended towards imposing EDI conventions on X.400 designers. Taken in their extreme form, the two views are incompatible, and can lead to serious conflicts when designing a protocol. However, it must be recognized that both views are valid as illustrated below: As the caption indicates, this model was colloquially known as the "cosmological model" during the work of the Associate Rapporteur Group. It turned out to be expedient to agree that the general conceptual model shown here, which embodies both views of the world, could be a useful guide when designing Pedi. When looking at things from the X.400 point of view, EDI appears to be a subset of X.400. When looking at things from the EDI point of view, X.400 appears to be a subset of EDI. In reality, as shown by the cosmological model, both X.400 and EDI are partly overlapping subsets of the real world, and neither is a subset of the other. Recognition of the fact that EDI is not a subset of X.400 led to the fundamental decision to accept for transmission in X.400 EDI interchanges encoded in character-based EDI syntaxes (like EDIFACT), instead of insisting that all EDI interchanges to be transmitted in X.400 be encoded in ASN.1. So you see a working group for international standardization can be an effective forum in which not only the interests, but the world views of different segments of the telecommunications community can be reconciled. In this case, everybody's perspective broadened, and the result was a standard, and a framework for future standards work, which was consistent with message handling system standards and with current EDI practice. The New ITU The restructuring of the International Telecommunication Union recommended by the High Level Committee was officially decided upon by a special Additional ITU Plenipotentiary Conference (APP-92) held in December 1992. The new structure includes a single Telecommunication Standardization Sector which consolidates the activities of the CCITT and the standards-setting activities of it's former radio counterpart, the CCIR. A simplified form of the new structure of ITU (omitting the Radiocommunications and Development Sectors and the Coordination Committee) is shown below. A new era began for the ITU on March 1st when the decisions taken last December by the APP entered provisionally into force, changing the structure, objectives and working methods of the Union. From that day the CCITT ceased to exist. It has been replaced by the Telecommunication Standardization Sector. That same day, the first World Telecommunication Standardization Conference (WTSC-93), opened here in Helsinki, taking the place of what would have been the Xth CCITT Plenary Assembly. Skeptics might think that these changes are a mere reshuffling of what existed so far while leaving the fundamentals untouched. This is not so. What happened on 1st March 1993 to ITU's standardization activities is not just taking down some doorplates and replacing them by others, but a decisive stage in the deep-rooted reform, started at the IXth Plenary Assembly in Melbourne in 1988, which will be carried on at WTSC-93 in Helsinki and beyond. Rapid evolution of telecommunication technologies and a dynamic, changing environment characterized by buzzwords like liberalization, privatization, competition, globalization and regionalization (to name just a few) are the forces shaping today's standardization. Only an organization which is flexible and adapts itself continuously to these determining forces will survive. The reform process started in Melbourne with an initial overhaul of the CCITT's structure, organization, documentation and working methods. One result was an acceleration of the production and publication of standards which, on average, now takes about 18 months compared to the four years or more before Melbourne! But this is no reason to rest on our laurels. Since Melbourne, both the technology and the telecommunication environment have changed and this should be reflected in the decisions of WTSC-93. In addition, the Helsinki conference will have to take account of the decisions of APP-92 concerning the transfer of network-related standardization from the former CCIR to the Standardization Sector and the creation of a Telecommunication Standardization Advisory Group (TSAG) as a new instrument to improve the efficiency of global standardization in the future. By the way, new recommendations and recommendations revised after 1st March 1993 will no longer be known as CCITT Recommendations. They will be designated ITU-T Recommendations. For example, when the revisions to CCITT Recommendation X.400 are approved by the WTSC, it will become ITU-T Recommendation X.400. Multi-party open EDI is an application which depends upon standards EDI is a particularly interesting example of a practical application which is built upon what's virtually a pyramid of standards. The standards for EDI For Administration, Commerce and Transport (EDIFACT) are a special case, in that there is significant user participation in a body which exists outside of the usual technical standards organs - specifically the UN's Economic Council for Europe - which is concerned with trade facilitation. Coordination with the traditional standards bodies, notably ISO and CCITT is handled with the Inter Agency edi Group (IAeG). The EDIFACT standard - which is also an ISO standard (9735) - deals with the application level syntax rules, that is to say, how the information related to specific trade documents or other well-specified forms, is coded. But EDIFACT doesn't deal with how the messages are carried and how a global network of users is addressed. In the series of standards known as Open Systems Interconnection (OSI), developed by CCITT in cooperation with the Joint Technical Committee on Information Technology, JTC1, of the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC), the communications process is treated as a seven layer model with standards defining services at each of the layers and the interfaces from layer to layer. The layers range from the physical layer up to the application level - which is where services such as X.400 Message Handling Systems (MHS) are situated. There is a whole family of standards known as MOTIS - Message oriented text interchange systems, for which there are parallel CCITT Recommendations and ISO/IEC Standards. The technical aspects of MHS are defined in the CCITT's X.400 series of Recommendations, developed by CCITT's Study Group VII. The system and service descriptions are found in the F.400 series of Recommendations developed by Study Group I. The overall system architecture of MHS is defined in Recommendation X.402. EDI, although it can be carried by many communication means, increasingly is being done with X.400. EDI contents can be carried in X.400 messages in a number of ways; common implementations either wrap the EDI data in an e-mail message - P2 - and then place it an X.400 envelope - P1, or, particularly in North America, they treat the EDI data as a P0 content and simply place it in the P1 X.400 envelope. These interim approaches were developed while awaiting the MHS recommendation for EDI and work with the 1984 version of the X.400 Recommendations. X.435 is based upon the 1988 version of X.400. The technical aspects of EDI messaging are defined in Recommendation X.435, approved by CCITT in March 1991. This recommendation formally defines the message handling application called EDI messaging (EDIMG). It defines a message content type and associated procedures, known as Pedi, tailored for the exchange of electronic data interchange (EDI). The corresponding service definitions are contained in Recommendation F.435. X.435 can be used with any content syntax for EDI, including EDIFACT or ANSI X12. X.435 is generally aligned with EDIFACT for consistency of definition of terms and data elements. EDI in a commercial setting is almost necessarily multi-partner. Greatest advantages accrue when systems can be used for a significant portion of transactions. Even though for many companies the decision to implement EDI is driven by a large customer's insistence that their suppliers communicate via EDI, suppliers also find benefits from rapid automatic processing of orders and shipping documents. In the government context, electronic submission of documents, using rigorously defined syntaxes assures uniformity which simplifies treatment of the documents. Customs processing of trade documents is a key area. This can provide savings for government, but even more compelling are the big advantages of speed and reduced clerical costs for shippers. There are obvious advantages of worldwide uniformity for this type of document, and much work is being done under the UN ECE in this area. Economy dictates the use of common infrastructures for messaging of all types as well as for a wide variety of other data communications requirements. This makes it natural that the EDI applications be atop a structure that is coherent, well-defined, and open. Open in the sense that multiple suppliers of services and equipment may be chosen without sacrificing the integrity of the entire system. In fact, Open-edi, rule-based exchange of information among the computer applications of independent parties, specifically includes commercial, technical, batch, interactive and all scenario-based EDI. The ISO/IEC Special Working Group on EDI (SWG-EDI) defines "open" from an EDI perspective as meaning: * "compatibility and harmonization with the OSI Reference Model * generic, public, interoperable, non-proprietary standards, testable with objective test criteria, conformance testing and certification * independent of specific business applications or industry sectors, fostering cross-industry sector edi utilizing existing ISO standards wherever possible and where necessary and/or justified using non-ISO standards as long as these further global interoperability in edi." The X.435 EDI application makes use of an X.400 system. Commercial or PTT operated public X.400 administrative domains (ADMDs) now operate in many countries. These ADMDs can be thought of as electronic postal services. A recent publication listed 60 different ADMDs in 36 countries! In addition, R&D X.400 networks, with connections to the Internet, exist in about 23 countries. Among suppliers of electronic mail software, not only is there virtually universal recognition of the need to provide an X.400 gateway, but there is increasing acceptance of X.400 even as "native mode" for many leading commercial e-mail software systems. Benefits of X.400/X.435 based EDI * Global network already in place * Security features * The standard that potential partners are likely to have * Network, protocol, equipment and staff shared with other e-mail * Diagnostic information for problem resolution * Widely available commercial software Choosing X.400 as the vehicle for EDI offers significant benefits (which are summarized in the list above). Operationally, this permits network and protocol consolidation (shared with non-EDI e- mail), requires fewer staff resources than an independent EDI network would require, and makes available the network management features of X.400 including a common tracking and auditing scheme with notification in case of non-delivery of a message, the ability to perform problem resolution and diagnosis, a mechanism for assuring non-repudiation, a choice of security functions such as authentication and encryption, common addressing scheme across many different application platforms, and access to X.500 directory resources. The number of external connections that an organization must maintain can be reduced dramatically by relying on an ADMD and the worldwide X.400 network to perform message transfer and routing. A number of international Value Added Networks (VANs) offer X.400 services, and interconnect with many of the national ADMDs. F.435 and X.435 include some useful features for EDI such as the notion of forwardable "EDI-responsibility," end-to-end acknowledgments, and security. X.400 messages may contain multiple parts, so that the same message could carry an EDI transaction, an accompanying text, and a binary file with CAD/CAM data. X.400, in turn, is an OSI application which runs over other OSI services including X.25 - typically the public switched packet data network (PSPDN). And the X.25 service, in turn, depends upon various elements of the telecommunications infrastructure (e.g., for transmission) which are specified in other standards. Thus the work of the standards bodies, national, regional, and especially international is relevant and necessary to getting to where members of a group such as yours can benefit from the use of EDI.