In navigating an aircraft visually over terrain the chance of successfully reaching the destination greatly depends upon the human element together with the visual aids provided. Some visual factors are studied in detail in this paper, particularly the detection and identification of ground features from the air. The suitability of various identification points for use as visual fixes is assessed quantitatively and the interaction between navigational accuracy and feature detection is discussed. The effects of visibility, speed and search are also studied. Visual aids considered include different scale maps, heading information, and fixed, zoom and search optics on television. A number of navigational tasks are covered, from following line features to direct cross-country tracks. Successive visual up-dating of a mixed manual and automatic navigation system is then assessed theoretically.

When studying the influence of navigational accuracy on collision risk, operating economy, or traffic organization it is necessary to distinguish between monitored and unmonitored traffic. With the monitored traffic the relative separation between aircraft is detected by the monitor and can be controlled by it. This is a closed-loop concept. With unmonitored traffic the separation is provided by procedural clearance of the flight plan. This is an open-loop concept.

The following discussion on the design of ships' bridges took place at a technical session of the Institute held in London on 6 January 1965. Captain Wacher illustrated the layout he proposed with a diagram.

Captain J. F. Wacher, P. & O.—Orient Lines. Existing general arrangements of the bridge, wheelhouse and chart room are inadequate for the needs of fast modern ships; the disposition of navigation, control, communication and safety instrumentation also leaves much to be desired. In recent years this has been recognized, prompted by an alarming increase in the number of ship collisions, particularly those which have occurred in restricted visibility.

Wing Commander Anderson has drawn attention to the fact that the distribution of the larger navigational errors is frequently observed to be non-gaussian in character. He raises the point whether this can be accounted for by the generation of such errors in succession, rather than simultaneously. The successive generation of navigation errors arises through the error incurred on a particular leg being distributed about a mean which is dependent on the error up to the end of the previous leg.

Loran C, a long-range hyperbolic navigation system, is currently used in transoceanic aircraft only to provide periodic position up-dating of primary self-contained navigation systems. The successful development and flight testing of an automatic coordinate converter has recently shown that the potential of the Loran system of navigation is vastly greater than past utilization has indicated. This paper briefly discusses the advantages of combining an increased capability Loran sub-system and an unsophisticated inertial sub-system within the framework provided by a heavy logistics support aircraft.

The surveillance of the air space by ground-based radar is important in air navigation and its military applications may be particularly mentioned since they have led to many important investigations, while aspects of the problem of less military significance have been neglected. Shore-based radar has also been applied to the surveillance of shipping in congested areas, captains and pilots being advised by VHF telephone though the captain retains full responsibility. This brings out a fundamental difference between the sea and the air. Air navigation and the ground control are integrated and the responsibility is shared. At sea the responsibility remains with the captain, and the use of radar increases this responsibility although ships are not yet obliged by law to be equipped with radar. Several court rulings have penalized the incorrect use of radar because of the greater responsibility that its availability entails and the recommendations of the Conference on Safety of Life at Sea (1960) concerning the Collision Regulations must also be remembered.

The safety of a vessel is the prime interest in ice conditions and routes must be selected which afford the least risk. However, this is only one aspect of the safe conduct of a voyage. The other is the expert handling of a vessel in ice-encumbered waters, or even when beset in ice. Both these aspects, the selection of the safest route and the expert handling of the vessel when ice is encountered, are discussed in this paper.

The paper describes an experimental analogue computer used with a civil marine radar set for determining the closest point of approach of selected ship's echoes and the time to closest approach. The equipment was installed on shore and measurements made of shipping passing to and from Portsmouth and Southampton harbours. The projected track of a ship echo was displayed on a 5-in. diameter cathode ray tube and the closest point of approach read as the distance of this track from the centre of the tube. The results show a high probability that under these conditions errors in C.P.A. will not exceed ±0·5 n.m. for targets between 10 and 6 n.m. range, and ±0·3 n.m. for targets nearer than 6 n.m.

Under the title Man or Machine? (Journal, 18, 10) a discussion took place in London on 25 April 1964, at a technical session of the Institute, on the extent to which automation could usefully be introduced into navigational procedures. Automation was taken to mean the use of machines which derive or accept navigational information and operate on it to provide a function for the control of craft. The discussion printed below took place at another technical session, held in London on 10 March 1965, when the purely economic aspects of the same problem were considered.

I should like to make some comment on Captain Wylie's article in the January issue of the Journal (18, 27). While not disputing his arithmetic I should like to state that in practice true motion is much more effective than he implies. For two years I used an Escort true motion set on the Southampton, Cherbourg, New York run and I can say most unequivocally that ships did not have to be proceeding at 18 knots to get satisfactory trails on the 12-mile range scale. In fact one could see useful trails when ships were proceeding reasonably slowly.