Today you can find volumes of information from any place in the world within seconds using Internet search engines. Students use this capability to find more information in minutes than previous generations could discover in weeks. As these students enter manufacturing companies as employees, they bring an expectation of the same search capability for corporate information. They expect that all information should be quickly searchable and available. Today you can find the menu, reviews, and a map to an unfamiliar restaurant within seconds, you can browse library archives on the other side of the world without leaving your desk, but in many companies you cannot find last month’s production numbers or equipment logs in less than a day of searching.

IT departments are addressing this information problem through enterprise search projects. Enterprise search describes applying search technologies, such as Google and Yahoo, to corporate data. Enterprise search differs from Internet searches in several fundamental ways.

Corporate data differs from publicly accessible internet data: it has numerous security and confidentiality concerns, the data is in many formats and databases, and often special applications are required to read the data.

Internet search users are looking for any answer and expect to see dozens or hundreds of documents, while enterprise search users are looking for the single right answer, often in a single document. Internet information is usually hyperlinked, allowing users to trace through connections to find information, while corporate data is rarely hyperlinked.

Internet data is frequently searched, so relevance can be determined based on access counts. Because there are fewer enterprise searches, relevance there must be determined by content. Corporate data also is often “dirty”; there are many misspellings, different terms for the same concept, department specific jargon, and the use of acronyms. A significant fraction of corporate data is in e-mail, but there are privacy and confidentiality concerns with making e-mail searchable. All of these differences make it difficult to find the single “right” answer to an enterprise search project.

It is important for Manufacturing IT to be ready for enterprise search when it rolls out. If your IT department does not yet have an enterprise search project, then they are probably investigating the technology.

Enterprise search is seen as a basic IT service, similar to databases and e-mail. An IT department may have a “search manager” that has responsibility for the corporate search servers and search application. The search service is managed by the IT department, but the content of the search is managed by the end user organizations. This is an important distinction, because it means that the content owners have some responsibility for enterprise search.

You should identify manufacturing IT “content managers” who have responsibility for an area, such as production or laboratory area. The content manager’s responsibility is to make sure that any changes to information location, all new systems, and all deleted systems are communicated to the search manager. The content manager should determine what files, databases, and directories are to be included in the enterprise search.

Typical searched manufacturing documents are batch records, validation documents, maintenance records, qualification records, test results, material information documents, equipment use records, sensor logs, and calibration logs. The content manager should also determine how frequently the information should be re-indexed by the search engine. Re-indexing is a network intensive task, so it is a good idea to setup separate directories and databases for long-term storage and for current information.

Meeting the expectations for enterprise search will require new roles and responsibilities in Manufacturing IT, so it’s important to be prepared for the coming wave of enterprise search projects.