 |
|
My first introduction to computers was a course in 1964 on FORTRAN IId on an IBM 1620. This was later applied to a physics research project which used laser light to measure precise distances. The experimental set up was as follows:
The FORTRAN program and its data had to be punched up on 80-column cards. It was then fed into the computer's card reader, compiled and run. Output was via a hammer-based typewriter.
For this work I designed and built (using experience from a previous fitting and machine-shop course in 1961) a special controlled atmosphere test apparatus. This I set up in the following experimental circuit to photograph the high-speed transients caused when the contacts broke the circuit.
As background for this work I read and greatly appreciated Max Planck's original book `Electromagnetic Theory'. This work was of great benefit when I moved to the Redifon Flight Simulator Division where I developed software for simulating the radiation patterns of aeronautical radio-navigation aids. Over fifteen years later, I developed more sophisticated software in C to simulate aerial radiation lobes.
While at the ERA I also helped to devise methods of measuring radio interference. This involved the use of many state-of-the-art radio receivers covering the whole of the then-used radio spectrum. One of my most fascinating experiences was listening to the whistlers - the glissando sound effects produced by the radio emissions from powerful lightening strikes echoing round the globe between the ground and the ionosphere.
I spent 2½ formative years writing programs for the Honeywell DDP 124 at the Redifon Flight Simulator Division in Crawley, Surrey.
It was a time of intense systems analysis and assembler programming in which I designed, developed and commissioned the simulation software for various types of aircraft navigation equipments and their terrestrial environments.
My whole time at Redifon was occupied on only two flight simulators. The first was the Douglas DC9-30 civil passenger aircraft for Flight Simulators Incorporated in California. The second was the Phantom F4-M fighter-bomber for the Royal Air Force [photo]. Photograph by Richard Cooke © Central Office of Information.
Below is an overview of these flight simulators' generic hardware:
Technical Achievements: During the time I was working on these two flight simulators I either originated or significantly enhanced the following:
I did most of the systems analysis, data structuring, programming plus all the acceptance commissioning for both the Radio Navigation & Communication systems and the Inertial Platform and Flight Director systems of both these flight simulators.
System Overview: Below is a generic system diagram of the radio navigation and communications simulation software on which I worked at Redifon.
GSD: A memory file of details such as lat, long, height, frequency, Morse ident, type (omni, VOR, TACAN, ILS, DME, marker) of the NAV and COM ground stations.
SCAN: a function which does a slow scan of all ground station data to shortlist stations within a zone where stations could be receivable.
LIST: Shortlist of stations which could possibly be receivable (ie a list of pointers to relevant GSD records).
R&B: Function to compute accurate range and bearing of stations on the shortlist using the spherical geometry formulae.
SIG: Function to scale each transmitter's signal strength according to its lobe radiation patterns, eg the horizontal lobe of an ILS localiser or the vertical lobe of a runway marker beacon.
TXs: Data array of all transmitters deemed to be receivable at aircraft's current location, output signals being keyed with station Morse idents.
TUNE: Function to attenuate signal strengths of transmitters according to the receiver's tuned frequency and selectivity pass-band.
O/P: Output of signal strength, ident, distance and bearing for NAV stations and voice for COM stations + graphical representation of these for flight instructor.
Technical Achievements: During the time I was working on this radio navigation and communications simulation software I either originated or significantly enhanced the following:
Below are the generic system diagrams of the flight director roll and pitch control channels simulation on which I worked at Redifon.
POS1: Aircraft Position: lat, long, height, drift, pitch, roll, azimuth angles and vertical accelerations.
INS: Inertial Navigation System simulation with built-in Earth-rate Drift and Transport Wander effects.
POS2: Positional Data: lat, long, height, drift, pitch, roll, azimuth.
CCA: Compass Compensator/Adapter: takes azimuth input from compass or directional gyro and applies mag var or earth-rate drift compensation.
HDG: Corrected true heading
FDC: Flight Director Computer azimuth channel: computes the roll correction which the pilot must make to bring the aircraft on to an asymptotic intercept with the pre-set heading or radio nav station radial.
POS: Aircraft Position: lat, long, height, drift, pitch, roll, azimuth angles and vertical accelerations.
ALT: Radio or pressure altimeter.
HGT: Radio or pressure height.
FDC: Flight Director Computer pitch channel: computes pitch correction which pilot must make to bring aircraft on to an asymptotic intercept with the pre-set height or ILS glide slope.
Technical Achievements: During the time I was working on this inertial platform and flight director simulation software I either originated or significantly enhanced the following:
In 1970, while I was employed by Scicon, London's Air Traffic Control Centre was being automated by means of a system known as `Mediator', the essence of which is shown below:
The part of Mediator to which I contributed was the flight plan processing system. Based on a triplicated Marconi Myriad 24bit/32k computer with two 6MB disks, this system:
Together with the annotated radar display, this provided the air traffic controller with the information he needed to advise and instruct the aircraft captain by radio. During my work on this project, I achieved two main objectives:
My first large assignment was to produce a detailed proposal for the upgrading of the TPC's automation from a simple camera copy production system based on IBM MTST composers to a computerised system covering the full document production task from creative writing to camera-ready copy. Using the most appropriate technology of the day I proposed a system, the essence of which is shown below:
In addition to the final design, I provided a phased implementation plan for the introduction to the new technology and the new working methods for the authors and camera-copy production staff. Having produced the proposal, my next task - which was as large again - was to produce a comprehensive presentation for an ITT Technical Directors' meeting in Brussels in order to market the idea to them to obtain the case funding required. This involved the production of a series of slides which could put over the essence of the system, the way it would be used, and the benefits to the ITT System as a whole in the more efficient production of better quality technical documentation. For the slides, I developed a somewhat unique illustrative style to convey this information quickly and effectively to the busy directors of this large multinational.
A complete representation using this diagramming method of the call processing system of the Metaconta 10C exchange was finally put together on a single diagram measuring 8 feet by 4 feet which became affectionately known as the Bayeux Tapestry. I also originated further diagrammatic methods and standards for depicting such things as the methods and techniques used for transferring control and data to and from routines on different interrupt regimes.
During 1975-76, I wrote a theoretical introduction, system description and user procedures for special traffic measurement features built into the software of the Metaconta 10C Toll Exchange. This model, which was based on stored program control cross-point technology, was built by Bell Telephone of Antwerp, Belgium to the specific requirements of the Australian Post Office. This task necessitated an exhaustive grasp of the theory and implementation of telephone traffic measurement and I made full and effective use of my mixed text + illustration style.
The main measurement was the occupancy of each subgroup of signalling devices - the exchange's main resources. This was measured for each sampling period throughout the day and presented as a histogram as shown. Exponential smoothing was applied to the successive samples to give a better feel for the daily variation.
Offered call rate and call dispersion were measured in a similar way giving the number of calls per route and the percentage dispersions of calls to down-route exchanges as illustrated in the example below:
Years later, I drew on the knowledge gained from this project when writing the traffic measurement program for an X25 network management suite.