Fullscreen HTML5Med Res (??MB)HTML4

THE WAY I SEE IT By NEVILLE WILLIAMS Police radar: where it works & where it doesn't Whether you know it or not, if you are booked for speeding by police using radar you have virtually no way of challenging it. This is despite serious reservations about its accuracy. When only one car is on the road, police radar is accurate. But if you are booked while travelling with other cars, you may not be the one at fault. You're driving along a two-lane country road in bright sunshine, chatting with your passengers and content with the world at large. The road is clearly sign-posted 80 km/h but you're consciously keeping up with the rest of the traffic at 90 kays or more, neither passing nor being passed. A police car looms up ahead, travelling in the opposite direction and you're silently thankful that he's going the other way. Of course, he could have had forward-looking radar but you dismiss the thought and go right on talking. But then you suddenly become aware of a wailing siren and there he is in your rear vision mirror, signalling you to pull over. You do so, wind down your window and, before your silent passengers, are officially advised that you were exceeding the speed limit when he passed you a couple of kilometres back. Even though everyone else appeared to be doing the same, you were undeniably exceeding the speed limit for that particular stretch of road. Not only that but you'd been indiscreet enough to leave a sufficient gap between you 78 SILICON CHIP and other cars travelling in the same direction to be exposed unambiguously to the radar beam. As for your friends, their reaction is typical: "Tough mate, yer gotta keep yer eyes open on that stretch" - not so much for road signs but for police! I didn't invent the above scenario; it happened to me a few months back, after 50-odd years on the road without a "bluey". Now, I am much more circumspect about keeping to the speed limit. It's quite difficult though, especially in traffic. Speed, a chronic problem If you doubt that statement, try sticking to 60km/h along ordinary suburban routes or the approaches to any country town. Most cars will pass you and disappear into the distance, many of them at a considerable rate of knots. Yes, I know the routine argument that the speed limits are unduly restrictive and that a good driver (is there any other kind?) will know when they can be exceeded without risk. And I know that the police are reputed not to worry too much about moderate infringements. Oh yes, and that a driver observing the limit is more of a hazard than one running with the traffic, because he/she "forces" others to pass. The simple fact is that speed limits apply nationwide and are meant to be observed as one means of curtailing our appalling road toll. They are abused, often flagrantly and dangerously because, in the absence of overt law enforcement, motorists "rationalise" their many and varied reasons for ignoring the speed limits and, in many cases, a few other safety measures as well: double lines and red lights, for example. Police radar Drivers dislike police radar because it poses a covert threat to their heavily rationalised disregard for the letter of the law. Why shouldn't they be able to add a modest 10 or 12km/h to the statutary limit if they're an experienced driver in a good vehicle, and/or the road is clear and dry, and/or they're running a bit late for whatever? Anyway, what guarantee do the police have that their radar is accurate? Haven't we all read about court cases where the charge was thrown out because the defence was able to cast doubt on the equipment, or the conditions under which the particular reading was obtained? Yes, I certainly have seen such reports and I've also noticed, at times, that the expert witnesses who've managed to discredit the ~~ ------'□ Ii.,..,______ RADAR (STATIC) I- TARGET R - - - - - - - (MOVING) Fig.1: When the distance between the radar equipment and the target is changing, Doppler effect produces a proportional shift in the apparent frequency of the reflected signal. The total path for the returned signal is 2R. , __ - tD ~--~~ ----=-& STATIC RADAR (a) I (b) . le FREQUENCY I I I (c) 14 13 ~ le l 11 I• 12 FREQUENCY ,. FREQUENCY Fig.2(a) shows a typical police radar set-up with a roadside testing station and four cars being scanned by the beam. Spectrogram (b) shows the signal frequency transmitted by the radar test set while (c) shows the signals received as reflections. Signal f1 is from vehicle 1, f2 is from vehicle 2 and so on. Spectrogram (d) shows the signals after they have been mixed, showing that the directional information is lost. This is why police radar accuracy is suspect when more than one vehicle is on the road. radar evidence in court have been elevated to the role of folk heroes, with other less fortunate drivers sharing vicariously in their victory over the police. The way I see it, on the balance of probability, most of the drivers picked up by radar will have been guilty anyway, despite alleged technical imperfections of the equipment or the way it is operated. It's fair enough that the accuracy of radar speed measurement should be challenged in contested situations but now that appears to be largely a thing of the past. If you are booked, that is virtually it. You front up and pay your fine or else! From the IREE "Monitor" Having made my own position clear, let's talk about an article published in the IREE (Institution of Radio & Electronics Engineers, Aust) "Monitor" for March 1989. Entitled "At Question The Overall Performance of Police Traffic Radar", it was written by Dr J. G. Lucas, of the Air Navigation Group, School of Electrical Engineering, University of Sydney. That Dr Lucas has strong reservations about radar-derived information is indicated in the introduction to the article where he refers to the fact that: "it was possible recently for the most sophisticated and up-to-date defence radar with its vast computer support on board a United States ship to mistake a commercial airliner for a small fighter/bomber with the ensuing disastrous consequences''. In the face of this, he continues: "It is perhaps not too hard to imagine that the traffic radar devices that are used on our roads can also make mistakes. Traffic radars ... are relatively unsophisticated devices' '. The final sentence in the introduction is the one around which the whole article revolves: However, legislation is being passed through Parliaments in Australia which decrees them to be infallible scientific instruments. As someone who has spent his whole career in electronics, I think that the idea of defining accuracy by legislative decree is ludicrous. Whether designing instruments or using them, accuracy has always been dependent on technical excellence - and is marked by progress along a never-ending learning curve. The measurement equipment we valued yesterday is questioned today and replaced tomorrow. That's the way it's always been. All states have, in fact, passed the legislation referred to by Dr Lucas so that now police radar is legally infallible. This defies precedents involving just about any kind of electronic equipment that was ever made - particularly if it involves a human input. Modern solid-state electronics can be good, frequently very good, and in some cases even superb. But infallible? In the minds of salespersons and politicians it may be but certainly not in the opinion of engineers. So what about radar, as used in JULY 1989 79 X-BAND RADAR CROSS SECTIONS 100 40.5 14.6 10 1.0 0. 1 0.1 .0 1 .005 .oo6T .00 1 ..,_ "' c:, :. "'z: ,_= = 31: "' cz: c:, c:,a:: ci!i > ::, ~~ 31: !l: z: cC w- "' ="' ,_ 35 = ,_ ,_ "" cc< c:,a.. 31: Ill z w "' I Cl z: cC ....► ....w ....w "' ,_ "' c:, c:, w ::, w "' a.. c., ► u cc ,_ c:, c:, :. cC Cl z c:, = 143.8 100 10 ... c., ::,- 10.2 1 1.0 0.16 0.1 0.16 .061 .01 .001 Ku-BAND RADAR CROSS SECTIONS Fig.3: using a Volkswagon Kombi Van as an arbitrary reference (1.0) for X-band and Ku-band radars respectively, this diagram illustrates the enormous difference in radar cross section between a Kenworth truck approaching (therefore front-on) and leaving, a Peugeot 505 and a motor bike. Australia? According to Dr Lucas, police radar typically involves a solid-state Gunn diode oscillator operating at X band (10.525GHz) or Ku band (21.15GHz) and producing some tens of milliwatts RF output, which is fed to a waveguide horn antenna. In the case of X-band equipment, the beam width is typically 20° between 6dB points, so that there is likely to be significant radiation over a 40° arc. In the Ku band, these figures are typically halved. In operation, the transmitted signal is directed towards the target vehicle. Some of the signal is then reflected from the target back into the horn for resolution by a zero IF (synchronous) receiving system (Fig .1 ). If the target vehicle is moving towards or away from the antenna, a Doppler shift is apparent in the frequency of the reflected signal amounting to approximately 20Hz per km/h at X band, and 45Hz per km/h in the Ku band. The return signal is higher in frequency for approaching vehicles, lower for those moving away. 80 SILICON CHIP will by the use of critically phased twin mixers - thereby alleviating that particular problem. This is the basis of the so-called "Slant" radar. In the mobile radars used in NSW, an IF sweep system searches for and locks on to fastest target. However, the system still can't distinguish between approaching and receding vehicles. In his opinion, the best scientific approach would be a procedure involving a Fourier Transform Algorithm which would provide accurate estimates of the speeds of all vehicles in the field of view. But, he says, "So far as is known, NO units throughout Australia use the Fourier Transform approach". The potential accuracy of speed measurement for an unambiguous target is typically well within 1 % . The catch in that statement lies in the phrase, "unambiguous target". As we shall see, an unambiguous target is not easy to obtain. In the midst of traffic, a simple stationary radar module is presented with multiple potential targets. It will sense multiple returns dispersed on either side of the transmit frequency, according to the speed of the vehicles and the direction in which they are travelling. This is shown in the spectrogram of Fig.2. This is confusing enough but, after mixing the transmitted and reflected signals, the directional information is lost because the resultants share a common IF spectrum. This means that the radar system cannot tell whether the measured speed is from an approaching vehicle or one that is moving away! Dr Lucas states that, in more advanced systems, it is possible to cancel the returns from either approaching or departing traffic at Radar "cross section" The other problem which prejudices the "infallibility" of speed radar is that of radar cross section. Some vehicles reflect a large amount of the radar signal while others reflect only a little. Many cars reflect less signal than motorbikes and some sports cars are practically invisible to radar, particularly if their pop-up headlights are down. Comprehensive tests, carried out at Sydney University, confirm this proposition. The tests involved driving typical vehicles towards X and Ku band traffic radars, with provision to photograph oscilloscope displays at the instant each approaching vehicle triggered an electronic gate. The results were plotted in terms of the vehicle's apparent radar cross section in square metres, as related to a hypothetical perfectly conducting sphere of equivalent cross sectional area. As will be apparent from Fig.3 (and I quote from the original paper) "there is NO simple relationship between the size of a vehicle and the return which it provides to a traffic radar device - it depends on the particular vehicle shape. "For example, the Peugeot 505 gives a smaller return than a motor bike. Some of the slick modern sports cars which are similarly virtually invisible become instantly significant when they raise their headlights". In the worst case, radar equip- MOBILE RADAR ment will tend to lock onto the vehicle with the largest radar return, irrespective of the direction in which it is travelling and without any direct indication to the operator as to the particular vehicle involved. In practice, a similar return could be expected from a Peugeot 505 approaching from 50 metres away, a large truck travelling 350-odd metres behind it, and the same large truck travelling in the opposite direction over 450 metres away! In mobile mode, Dr Lucas concedes that typical equipment can give an accurate reading when the two vehicles are alone on the particular stretch of road. The equipment assumes that the fastest vehicle return will be the converging speed with the target vehicle, from which it must subtract the speed of the patrol vehicle. However, he says, the devices will continue to give an ostensibly "valid" reading when there is more than one vehicle in the field of view. Continuing: "In the worst case, the operator will continue to get a reading when there is traffic moving in the same direction as the patrol vehicle". Why? I don't quite follow unless he has in mind a large truck several hundred yards ahead of the patrol vehicle really burning up the bitumen! The point he does stress, in closing, is that the operator of a radar patrol vehicle has precious little time to make checks and decisions and the opportunity only to make the briefest of notes about the details of any one incident. His conclusion: "All traffic devices do operate extremely accurately when there is only a SINGLE target vehicle on the roadway. In ANY instance where there is more than ONE target vehicle, there is ALWAYS doubt as to which vehicle the radar has acquired". Fig. 4: in mobile mode, the radar assumes that the highest reading is the converging speed with the target vehicle, from which it must subtract the speed of the patrol vehicle. The way I see it While I have not studied the subject in detail, I would be surprised if Dr Lucas' article says anything that has not been said at other times as, for example, in anti-radar court proceedings which have been reported in the media. What it does do is to restate the case against existing traffic radar systems in an organised and relatively unemotional way. In the face of such doubts, it is ludicrous that Australian Governments have decreed such devices to be an "Infallible Scientific Instrument". Courts are now presumably obliged to accept radar evidence and reject defence submissions seeking to discredit it. As I said at the outset, my motive in publicising Dr Lucas' paper is not to see traffic radar totally discredited. As one means of enforcing speed limits, it has a contr.ibution to make to saving lives on the road. But to impose by law a fictional "infallibility" which cannot be supported scientifically would be a clear miscarriage of justice. The end certainly does not justify the means. Defence must be preserved as an option for those who choose to plead "not guilty". And if, as a result, traffic radar needs to be rendered more credible by improved technology, by automated realtime documentation or by photographic evidence, so be it. Reference If you want to study Dr Lucas' article in detail, you can get a copy from the IREE Head Office. (Commercial Unit 3, 2 McLean St, Edgecliff, NSW, 2027. Enclose $2.00 to cover photostats and postage). ~ MAKE YOUR DREAMS COME TRUE LAPTOPS FROM s1995.oo * INTERNAL BATTERY, EXTERNAL AC ADAPTOR *LIGHT, FAST, 10MHz *COMPACT AND CONVENIENT * SUPER TWIST LCD DISPLAY *64OKB RAM *SERIAL AND PARALLEL PORT * REAL TIME CLOCK LASER PRINTER s2995.oo *6 PAGES PER MINUTE *SERIAL AND PARALLEL (DUAL HOST) * 6 BUil T-IN FONTS * HP LASER JET SERIES II EMULATION * OPTIONAL IBM PROPRINTER EMULATION * 4 ADDITIONAL SOFT FONTS INCLUDED * EXPAN DABLE TO 4.5MB COMPUTER CARE KIT s115.oo * 6 CLEANING KITS IN ONE *SUITABLE FOR ALL COMPUTERS _;<t··::..._,,,_...,,,.....,, - - - - -- zz:. •! ,,l-~ ALL PRICES INCLUDE SALES TAX FREE ALL PURPOSE CLEANING KIT WITH HARDWARE PURCHASES FREEPOST NO. 31 ESCKAY PRINTERNET PO BOX 2591, NTH PARRAMATTA, 2151 TEL: (02) 630 8956 JULY 1989 81