Title: Designing for man in an automated world
Pages: 58 - 63
Author: Ronald E.Fischbacher
Text: Designing for man in an automated world
by Ronald E. Fischbacher
The fiftieth anniversary this year of the Scientific Instrument Makers' Association is being supported by an exhibition of modern British instruments at The Design Centre from October 5-November 19. In the article here, the deputy director of the Scientific Instrument Research Association describes some of the major technical developments in the fast expanding field of industrial measuring and control equipment, and goes on to discuss some problems that these developments pose for the designer. His article is followed by three short case histories of instruments included in the exhibition.
It is almost as difficult a task to define an instrument as it is to define design: the definitions of both have been argued at length. Even to limit the scope to scientific instruments leaves a very heterogeneous field. In this article, I shall restrict myself to the field of industrial measuring and control equipment used in the process industries. This is an important, fast growing, fast changing field, worthy of the attention of the industrial designer. For a subject so large, this can be no more than a brief introduction.
The opportunity - and the need - that exists for the industrial designer can readily be illustrated. When I set out to write this article, I was surprised to encounter some difficulty in summoning up a notable example of a well designed instrument in this field. Suspecting a mental lapse, I posed the question independently to two prominent instrument users. The reply was the same in both cases: "Certainly there must be some, but I can't think of one just at the moment". The conclusion is inescapable - that there are few classic designs in this particular field.
Automation and miniaturisation
Before looking at the factors directly affecting design, we must consider the forces at work in the development of new instruments. First is the trend to automatic rather than human control. Whereas in the past the prime purpose of an instrument was to convey information to a man, now it is more likely to be required to feed information to a computer, controller, data logger or actuator. It may also have to impart its information to man, but this is becoming a subsidiary role.
The second important trend is that of miniaturisation. This is not just a current fetish. Increasing use of instrumentation on large plant installations means that either control rooms become impossibly large, or instruments become ever smaller. To an appreciable extent, too, printed circuits, miniature components, and now possibly integrated circuits, contribute to the reliability of the modern electronic instrument.
A third constraint on the designer is i m posed by the modular form that the panel-mounted units of instrumentation systems are assuming. The lack of a single standard size for these modules is an immediate difficulty that will no doubt be resolved in due course.
An instrument may perform some or all of the functions shown in on the following page. These functions may all be included within a single unit, or there may be a separate unit performing each function. For example, a turbine flowmeter may comprise a turbine housing in series with the plant flow line. a display in the control room, and an associated electronic unit at some intermediate point. Alternatively, the output from the electronic unit may be fed into a computing unit, along with the output from a density meter, to obtain mass flow. What then constitutes the instrument ? In this light we have to consider the design problems of the separate elements or components individually.
Design of the transducer (the primary sensor) is to a very large extent dictated by the conditions under which it will have to work, and transducers have to work under the most arduous conditions imaginable -from the heat of the blast furnace to the corrosive power of liquid sodium. These tax the ingenuity and ability of the materials scientist and the mechanical engineer.
Because of this, the contribution which the industrial designer can make to economy of design, ease of accessibility and ready replacement may be overlooked. Robustness rather than aesthetic appeal is
Designing for man in an automated world
1. The basic functions of an instrument The transducer detects the measured quantity and transforms it into a suitable form (electrical, mechanical, pneumatic, etc) for further processing. A microphone, for example, translates sound vibrations into electrical impulses.
The processing unit modifies the signal in such a way as to enable it to be used to display information, or to enable it to be used to actuate or control. The most usual function of the processor is to amplify, as in the case of the microphone. It may, however; perform much more complex operations in the form of analogue or digital computers.
The display may take a number of forms - digital (eg, numerical display tubes or typewriter print-out), analogue (eg, meters or pen recorders), or a combination of both presented on a cathode ray tube. The sound coming over a loudspeaker may be regarded as an aural display.
The actuator is a control device operated by the signal, and may take the form of a switch (eg, in a temperature controller) or a valve (eg, in a flow controller). It is normally part of the control system, but in some cases may be part of the instrument itself.
2. This instrument controls temperature at a pre-set level. The knob in the bottom left hand corner sets the lower pointer to the desired temperature, and is removable to prevent interference. Maker Ether Controls Ltd. a desirable feature in a piece of equipment which must withstand being clambered over, hosed down and belaboured in transit by a variety of solid objects. The designer can therefore contribute to the durability of the external housing of the transducer, to its compatibility with environment, and to the protection afforded to the often delicate mechanism or elements within.
But he meddles with that mechanism at his peril; the constraints are too often already those of the limits of science or technology and manufacturing skill, rather than those of ergonomics or aesthetics. Nevertheless, the originator needs to be exposed to the pressure of the industrial designer to be certain that an optical system cannot be 'folded', or have its components rearranged, in a manner which will facilitate the design of a more compact or acceptable unit without detriment to the performance of the transducer.
Processing or operational units
The equipment may be electrical, pneumatic or hydraulic, but since the former is likely to predominate we shall use it as an example. Similar arguments are likely to apply to the others.
We have already noted that the most marked trend is to miniaturisation. Present generation electronic units employing transistors and printed circuits make possible volume reduction by a factor of 10 or more on comparable valve and conventional wired-chassis construction. Future generations of equipment employing thin film or integrated circuits foreshadow further volume reduction by factors of 100, or even 1,000, in operational circuitry, 8. Similar reductions are to be expected in pneumatic logic devices and fluid amplifiers.
But the size of human fingers and the resolution of the human eye remain the same. Some of the photographs illustrate the problems that already arise in designing instrument units. Whatever the future may hold by way of computer control and data processing, the day is still far off when instruments will not have to be set up by hand, or when reversion to manual control in emergency situations will not be necessary.
For some time to come, therefore, we shall have to face up to the problems posed by the necessity to accommodate control knobs, power and signal connectors and local indicators or meters, on panels which are steadily shrinking in size. Equipment whose primary function is to control without human intervention must still be compatible with human manipulative and interpretative ability. Here indeed is a challenge to the industrial designer; too many of the instrument panels to be seen today are gestures of despair on the part of the engineer who has not appreciated the contribution that can be made by the trained designer. Fresh thinking must be brought to bear on this problem, which is likely to become yet more severe.
A further problem arising from this era of transition between human and fully automatic control is illustrated in I. The fear of the unauthorised 'knob twiddler' has prompted many solutions, none
3-5 The unit, a, from a modular amplifying system, illustrates the problem posed by the need to accommodate a large number of control knobs on small modular panels. The crowding becomes even more marked when the modules are mounted in a rack assembly, 4, and emphasises the problems both of identification and physical control. The latest version of the instrument, 5, shows how careful design can make a considerable improvement. Maker S. E. Laboratories Ltd.
entirely satisfactory, to the problem of making authorised adjustments easy, but unauthorised adjustments difficult or impossible. The knob in the lower left hand corner is used to set the point at which this recording controller controls the process. The knob is removable so that it can be retained by the person authorised to make the adjustment; but a removable knob or control is one which can also be lost. Many alternative solutions (such as lockable cover plates) have been tried, but it is difficult to find an entirely satisfactory one.
Maintenance depends very much on the accessibility of sub-units or individual components. Accessibility in general has improved enormously over the last decade or so, but in many instruments component packing density will inevitably remain high; and to ensure ready access to each component requires careful attention to detail. Talk of 'throw away' replaceable units has long been heard, but such a policy is as yet far from being economical.
Once integrated circuits or miniature fluid logic elements are in widespread use, it will then become necessary to throw away faulty or damaged sub-units, since this type of operational element is not likely to be repairable. Today, the largest users of instruments may find it possible to maintain a stock of spare sub-units for instant replacement, allowing time to repair faults at leisure or to return the defective unit to the manufacturer; the small user certainly cannot afford the luxury of ready spares.
It must be borne in mind that ready removal of parts does not alone ensure good maintainability in this context. To be able to remedy faults, it is usually necessary to have access to the equipment when it is working. One approach is to give effective accessibility by providing 'patching' connections to a sub-unit physically removed from the main unit.
The ergonomic factors in i nstrument units do not differ essentially from principles already well established and frequently described, once it is realised that the ancillary role of manual operation is nevertheless a vital one. Much has been written about informative, unambiguous labelling; but these well established principles are still not being applied as widely as they should be.
Keeping displays readable
Two forms of display may be regarded as traditional in the process control field. The first is the moving pointer meter, about which a great deal has been written, and which has undergone many improvements in recent years. The second is the chart recorder, with circular or strip chart as the recording medium.
Improved readability of pointers and scales has resulted from practices now codified in BS 3693. Improved bezel design has at the same time improved scale illumination. Once again, however, the trend to miniaturisation, especially of instrument-mounted meters, and the secondary role of many meters which have to be read only in emergency or start-up situations, can lead the unwary to install a meter of a size which taxes or defies the eye of the user, although it
6. A further aspect of the problem of accommodating controls and terminals on small modular panels is shown in these units. The cover plate shown in position only on the left hand unit protects the connections and helps to reduce clutter, although when the panels are connected the leads emerging from behind the cover plates will still make it difficult to keep the instrument looking tidy. A group of units in this range is also illustrated on page 58. Maker Bailey Meters & Controls Ltd.
looks neat on the panel.
Some of the functions of pointer meters are being replaced by digital displays. A great variety of these exist, some ill adapted to quick and accurate reading, others inadequately visible under widely varying conditions of ambient illumination or viewing angle. Care must be exercised not only in selecting the most suitable form of digital display, but also in deciding whether in the circumstances digital or analogue display is most appropriate. Digital display can be as accurate as the associated equipment allows (but no more so), while analogue pointer information can be more instantaneously absorbed.
Chart recorders have also shrunk in size, to the point where it is claimed that the requirements of accuracy have to take into account the ratio of chart width to inkline width. An increasing amount of data is, of course, being stored digitally in computer-compatible form. There are many situations, however, where the analogue trace on a recorder can give speedier understanding of trends and the nature of change. In such situations, the ancillary role of the recorder must not be allowed to overshadow its ergonomic functions.
Mimic diagrams and display consoles also pose problems that are not always fully appreciated. The pictorial nature of a graphic display can readily distract attention from the fact that its most frequent function is to indicate promptly the occurrence of 'off-normal' conditions. Since the human eye readily detects lack of symmetry or uniformity, 'normal' conditions can be made to coincide with aligned pointers, uniformity of coloured indicators and other devices which eliminate the necessity of checking actual readings.
What of future trends ? In the mainstream stands the computer analogue, digital or hybrid. The item which is likely to stand out longest against the computer is the transducer. It is likely, however, that the transducer will incorporate within it, or close to it, microminiature circuits which can be interrogated by the computer. Processing units may then largely disappear, being replaced in function by computer sub-programmes. Displays in present form are likely to be subordinated to a master display fed from the computer. Such displays are already in use in advanced applications. Desired information can be called up by simple selection by touch or by light-pen from a list of electronically presented alternatives on the screen.
Such fundamental changes in concept are already within reach; yet they represent only one end of a spectrum of items of plant instrumentation that extends from control of a complete oil refinery down to a single temperature controller in a small process shop, where any computer is beyond reach. As far ahead as we can see, man-machine compatibility must be maintained at the component level as well as the system level, and the most exciting developments which are now taking place in the industry must not be allowed to obscure this basic fact.
7. This unit, introduced at this year's Instruments, Electronics and Automation exhibition, illustrates developments which are taking place in centralised displays. Information can be called up from a computer or data store by touching the appropriate point on the screen. Similar systems using a light pen to call up new data are employed in other systems such as the American Sketchpad computer aided design equipment (DESIGN 213/47). Maker The Plessey Co Ltd.
8. The group of five valve circuits at the top of this picture, the transistorised circuit, bottom led, and the microcircuit next to it, all perform approximately the same function. Although these circuits have been developed for computers curing the past 10 years, the size reduction is indicative of similar changes which are taking place in the whole field of scientific instruments. Maker Elliott Automation Ltd.