Mnematron

The science fiction author Arthur C. Clarke created many very important new concepts in science, communications and information. He is credited with the concept of the geostationary satellite which is now a key element in global communications. In creating a new concept, he often had to create a neologism to name the entity that was to deliver the functionality described by the new concept.

One of his least known contributions was to create the name mneumatron to describe an intelligent information repository. In the early 1970s, the USAF decided to use this name to describe a classified information project which was intended to produce intelligent audio-visual filing cabinets.

The limitations of hardware technology and communications bandwidth, at that time, prevented the project from producing more than a concept demonstrator, as a starting point for further work. The first systems used mainframe computers and stored Multi-media data on magnetic drums. Access was mainly from computer terminals directly wired to the system, although some experiments involved remote access by terminals connected through the low-speed (typically 300 bit/s) modems of the time, over analog telephone cables. The very high cost of development meant that this technology was to be confined to large and costly mission-critical systems employed by military and intelligence agencies.

The first systems held data in text, audio and video formats, but the level of Artificial Intelligence was very limited. As concept demonstrators, these pioneering systems showed that computer and communications technologies could be profitably converged and that Artificial Intelligence, AI, could be employed to automate the process of retrieving data that was stored magnetically in many very different formats.

The classified nature of this, and following projects, meant that the mneumatron was known only to small groups of individuals who were working on the development of advanced command and control systems. The mneumatron concept became the heart of the advanced Command and Control System known usually as C³I (Command, Control, Communications, Intelligence) or C⁴I² (Command, Control, Communications, Computers, Information, Intelligence)] . The mneumatron environment was produced in several architectural designs by different teams of engineers and systems architects. In many cases, the mneumatron environment was part of a Trusted Computer System and this became increasingly necessary as these systems were networked.

One technology that made further mneumatron development practical for classified information systems was the Compartmented Mode Workstation, CMW, and its Trusted Networking (DINSiX/MaxSix) environment, developed to specifications created by the US Defense Intelligence Agency, DIA. This allowed mneumatrons to hold data at many classification levels, each in a trusted compartment, or virtual computer, that could be remotely accessed by any user who held the necessary Security Profile and associated privileges. This approach reduced the probability that an intelligence analyst would be denied access to information necessary to that analyst's duties.

In modern intelligence environments there are a number of potentially constraining factors. The first is, frequently, an information overload, where information is pouring in from many different sources, and where the major risk is that vital information will be buried under mountains of trivial data. The second risk is that an analyst will be unable to access vital information because it has been acquired by another department, or agency, that has protected, what it sees as, very sensitive data, unaware that it may be urgently needed by another group. The third constraint is caused by the risk posed through data aggregation.

This data aggregation factor is often under-appreciated and few fully understand this danger that affects information in private, commercial and governmental information systems. Many items of data made be stored and processed in a computer-based information system. Each item has a low sensitivity on its own. However, as a series of related items of data are grouped together by the computer, the combined group of data items may represent highly sensitive information that must be carefully protected.

Artificial Intelligence, developed under mneumatron and CMW projects, has allowed a much greatly flexibility of access to data that represents a wide range of risks and values. This improves the quality of analysis and speeds the production of vital reports. It also makes possible the presentation of the most important information in real-time, with the immediate option to view related information to aid fast and accurate decision making.

By the 1990s, the basic mneumatron architecture was employed in a number of Emergency Operations Centers, EOC, that were manned by police, fire and rescue personnel to deal with major and extreme civil emergencies. The functionality was very similar to military and intelligence C4I2 systems of the time, but scaled down and mounted on lower cost hardware. Security was also handled differently on the assumption that it was less important in EOC systems - often a point of contention during design between the police, who wanted security, and the fire/rescue personnel, who wanted fast access without constraints. Subsequently, some EOC systems have greatly increased the methods of protecting data, as they replace systems with the next generation, having better appreciated the growing risk issues of widely networked systems.

In 1995, the mneumatron was selected to form the core of the FIRE Project, a community service project intended to provide knowledge from on-line public access information resources. This project uses Content Management System, CMS, and Weblog portals as the front ends to the mneumatron. Typical portals are Broadly Boats [http://bb.firetrench.com] and FIRE News [http://ftnews.firetrench.com] that provide breaking news columns at portal level and provide access to the growing information resource held within the mneumatron.