Computer-aided Paper Chart Production in Japan

The Japan Hydrographic Department began developing a computer-aided production process about twenty years ago, but this proved difficult due to limitations in computer technology. Subsequent progress in this area has helped development of a computer-aided chart compilation method, resulting in unimagined improvements in efficiency. This article describes events of the past twenty years.

Because of their indispensability for maritime navigation it is important that navigational charts be produced within as short a time period as possible and that their contents be constantly updated with the latest information. Despite this, the fact of most chart production processes being carried out by hand meant in the past an enormous volume of manpower and man-days. The Hydrographic and Oceanographic Department of Japan Coast Guard (JHOD, then Hydrographic Department of the Japan Maritime Safety Agency) therefore began towards the end of the 1970s to research the possibility of producing paper charts by computer.

First Steps
A chart consists of several colours, each colour being assigned to some kind of information such as depth, depth contour, light buoy, traffic route, shoals and land. Each colour was at that time duplicated onto film, then assumed to be the original plate, and subsequently onto an aluminium machine plate. The last step was printing as a paper-chart in the printing factory. At first, digital processing was adopted only for radio navigation lattice and for the skeleton sheet, because of their simplicity. In 1984 equipment such as the digitiser was introduced to pick up sheet coordinates, and the plotter to make negative film for the ‘scribe’ method in which opaque coatings were shaved off the film. The result was publication of lattice charts for Decca and Loran A. Drawing software was also improved, and then the first computer-made chart, No.603 ‘Bengkulu to Selat Sunda’ was published in 1986. This chart was selected because the geographical names were all expressed alphabetically and all coastlines by single lines. Another reason for picking this chart was the paucity of data involved. For the production of a land tint-plate to express land area, a cutter was installed on the head of the plotter that plotted coastlines and graticules on the film upon which was pasted a thin opaque coating; coatings were then peeled off to express the land.

First Evaluation
A similar method was used for two more chart publications. At the time this method achieved an astonishing shortening of working time, from two weeks to several hours, with a skilled drafting engineer. However, it did not satisfy so many chart compilers because the scribe-plotter was unable to express Japanese characters with a lot of strokes, or a complex symbol. Nor could it express coastlines as a painting. As a result of re-examination, photo-composition was reintroduced for characters other than digits, latitude and longitude of chart limits and depths, and a full computer mapping system followed one step behind. A major turning point had arrived for both the system and the software. Standards and the specifications for the electronic chart had begun to emerge in the atmosphere surrounding the International Hydrographic Organization; a navigation system using GPS and the electronic chart was expected to become reality.

Towards the EC
This keen interest came in Japan too to be turned towards the production of an electronic chart that pursued safety and efficiency of navigation. Specialists in electronic charting from every advanced country were invited to a seminar on the production of the electronic chart held at the Japan Hydrographic Department in 1989. This seminar encouraged electronic chart development in Japan, and the work of digitising paper-charts for an electronic chart based on international standards was planned. The introduction of a system for the production of electronic charts began in 1994. The photo-head plotter, a so-called second-generation system in which a film was exposed to light in order to make an image of it, was also introduced. This system was used to concurrently produce both electronic and paper charts. At first production levels of complete paper charts using this system were not high; the principal target had been production of electronic charts. Paper charts were derived from the electronic charts, an unexampled method at that time and one with which it took more time to compile a paper chart than expected because of heavy overburdening of the software.
Containing a lot of information useless for paper chart compilation, electronic chart data caused a problem of complicated work and software function for paper chart compilation. Another problem was that electronic chart data was restricted by the ENC cell boundaries. An electronic chart of Japan did not have the same coverage of the corresponding paper chart, so that one chart was divided into several cells. A cell of the electronic chart reflects an overlapping of one paper chart upon its neighbouring ones; the corresponding paper charts needed digitising independently of electronic chart data.

Development Problems
Symbols and expressions are not always the same in electronic charts and paper charts. Therefore a conversion table was prepared in order to fix line thickness and symbols on the paper chart from the corresponding code in the electronic chart. More than six hundred kinds of analogue symbols from paper charts were greatly extended in the photograph, read with the scanner and converted into digital symbols. Furthermore, inclined and simplified symbols had to be prepared which were not listed in ‘chart symbols and abbreviations’. There was also a problem in the kinds of characters; those that had been traditionally used for the paper chart became untenable in the new system, so that extremely similar kinds of characters had to be adopted. Each kind of character had a conspicuous design, differing width and different shapes in the same font size. This often made it impossible to put the new characters in the same place as before.
In addition, expansion and contraction of the new film was greater because the new film was thinner than the one traditionally used for the output of the photo-head plotter. Much time was spent in establishing a method of securing accuracy against expansion and contraction whilst keeping the temperature and humidity stable until the film became steady. Most time-consuming in this process was arriving at the understanding that expansion and contraction were greatly and especially influenced by humidity. After many trials and errors, the first chart under this system, No.81 ‘Tori-Sima to O-Shima’ was published in 1995. The chart was satisfactory in terms of characters and symbols and, even more important, in no way inferior to the one produced by a traditional drafting engineer.

The Next Step
Although computerisation in the chart production process was a long-term project for the Hydrographic Department, it faced short-term problems of capability originating in a shortage of CPU/software for compiling maps with large volumes of information. There was also the fact that there was no output device to express complex figures in the first-generation system. The second-generation system solved most of these problems; greatly improved processing performance came with rapid progress in computer technology. However, this did not mean that productivity rose accordingly. Delays were experienced in filling the database and there was lack of training for computer-system editing staff. To resolve this, digitising of the paper chart was started independently of digitising of the electronic chart, which would require a great amount of time and cost, and computer-system editing staff were switched only gradually from the traditional method.

New Paper Charts
It took more than twenty years after initial aims at automated chart production to produce charts by computer that were not inferior to the ones made by the skilled engineer, and that were also of consistent quality. Today the computer-aided system has become editor-friendly thanks to many meetings for exchange of ideas between software developers and skilled engineers. Working icons proposed by an engineer graduated in psychology allowed editing staff easily to remember and get used to them. As a result, the present productivity of one person has increased five-fold or more compared to his productivity at the time system development began.
Computerised equipment for hydrographic surveys has brought in digital survey results, which has in turn made possible ‘chartlets’ for update and production of new editions of the paper chart under the system. Mutual data exchange between electronic and paper chart has, moreover, promoted the development of a database and targets are now being set for completion of the database for all navigational charts published by JHOD. Some experiments are being advanced in order that a remote user may print an updated paper chart out of the database via the computer network and that a printing factory may directly produce an original print plate without production of the film. "Was all this produced with the computer system?" asked one of the retired engineers when he saw the latest chart at a reunion.