79. Evidence of speed
(1) If in any criminal proceedings the speed at which a motor vehicle or trailer travelled on any occasion is relevant, evidence of the speed of the motor vehicle or trailer as indicated or determined on that occasion by a prescribed road safety camera or prescribed speed detector when tested, sealed and used in the prescribed manner is, without prejudice to any other mode of proof and in the absence of evidence to the contrary, proof of the speed of the motor vehicle or trailer on that occasion.
The prescribed devices and testing, sealing and use requirements are contained in Part 3 of the Road Safety (General) Regulations 2009.
The result of this provision is that if the police use a radar or laser speed measuring device to accuse a person of speeding, the reading on the device is accepted as the driver's actual speed unless the driver adduces ‘evidence to the contrary’ to overcome the prima facie effect of the legislation.
The South Australian Supreme Court recently handed down a judgment showing how that might occur.
In Police v Hicks  SASC 136, Mr Hicks was charged with speeding at 81 kph in a 60 kph zone. The police used a laser, which enjoys a similar statutory presumption to those in Victoria. The Traffic Act 1961 (SA) s 175 relevantly provides:
(3) In proceedings for an offence against this Act—
(ba) a document produced by the prosecution and purporting to be signed by the Commissioner of Police, or by any other police officer of or above the rank of inspector, and purporting to certify that a specified traffic speed analyser had been tested on a specified day and was shown by the test to be accurate to the extent indicated in the document constitutes, in the absence of proof to the contrary, proof of the facts certified and that the traffic speed analyser was accurate to that extent on the day on which it was so tested and, for the purpose of measuring the speed of any motor vehicle—
(i) in the case of a traffic speed analyser that was, at the time of measurement, mounted in a fixed housing—during the period of 27 days immediately following that day; or
(ii) in any other case—on the day following that day, whether or not the speed measured differed from the speed in relation to which the analyser was tested or the circumstances of the measurement differed in any other respect from the circumstances of the test;
The accused driver said in evidence he wasn't speeding:
 The respondent gave evidence in his defence and called his wife; she was a passenger in the vehicle at the relevant time. Both gave evidence that the vehicle was not travelling as fast as 81 kilometres per hour at that time. The respondent himself said that shortly after leaving the roundabout he saw Senior Constable Turner and saw his flashing lights go on. He immediately looked down at his speedometer and saw that he was travelling at 50 kilometres per hour...
The Magistrate dismissed the charge. The police appealed to the Supreme Court.
The Supreme Court dismissed the appeal, noting at  that the accused driver didn't dispute the accuracy of the device but instead relied on the result of providing ‘evidence to the contrary’ of the speed shown by the laser.
I'm actually a little surprised at this result. Although the legal analysis is sound, from the facts as they appear in the judgment it seems the driver checked his speed some time after the police officer pinged him. This is the common problem for most drivers who want to rebut the prima facie provision, because they're often not able to swear precisely what their speed was at the exact time and place the police saw them. And it's often compounded by concealed or unmarked police cars.
There's quite a few cases that make this point. In Madgen v Ashe (1992) 17 MVR 219 the West Australian Supreme Court said evidence from a driver that he wasn't speeding wasn't competent or cogent evidence that displaced the prima facie evidence of the prosecution cases. There, the driver admitted to only estimating his speed without looking at his speedo, and was wrong about his speed and location estimates.
That case was cited in Davis v Armstrong (1993) 17 MVR 190, another West Australian case that explained the point neatly at 192:
The burden then fell upon the applicant to adduce evidence capable of displacing the prima facie effect of the prosecution evidence: Madgen v Ashe (1992) 17 MVR 218. In other words the onus was thrown upon the applicant to prove on the balance of probabilities that his speed did not exceed the limit at the relevant time. He could do that by displacing the prima facie evidentiary effect of the speed measuring device by throwing doubt on its accuracy by establishing that it had not been properly tested, or that the circumstances were such that it might not have operated accurately exclusively with respect to the applicant's vehicle, or he could seek to displace the prima facie evidentiary effect of the speed measuring equipment by persuading the court of trial to accept evidence, either his evidence or that of other witnesses, to the effect that he was not exceeding the speed limit: Cazzol v Fuss (1988) 6 MVR 350.
In that case, the driver lost his appeal because he challenged the accuracy of the speed camera rather than trying to displace the prima facie effect of the evidence. Cazzol v Fuss (1988) 6 MVR 350, cited in that case, succictly made the point at 352 about overcoming the prima facie provision:
[A] defendant and other witnesses may give evidence on oath that they are able to assert positively that the speed of the vehicle was 60 km/h or less. If that evidence is accepted then the defendant would have discharge the onus that has been placed upon him.
Perkins v Pohla-Murray (1983) 1 MVR 165 and Hizaji v Orr (1997) 26 MVR 266 are ACT judgments along the same vein.
The difficulty for a court determining these contests is when a credible witness gives cogent evidence to say "I wasn't speeding" (or, "I wasn't going that fast") and the prosecution adduces cogent evidence to say the accused was speeding. What then?
I haven't found any cases precisely on this point, but I think it's resolved the same way as any conflict between competing events, applying a Liberato direction from Liberato v The Queen (1985) 159 CLR 507 (and also R v Calides (1983) 34 SASR 355). The Judicial College's Criminal Charge Book provides a nice summary of this at 1.7.1 - Bench Notes: Onus and Standard of Proof:
76. While it is not necessary to give a Liberato direction in every case where the jury is invited to decide whether the prosecution witnesses or the defence witnesses should be believed, it is desirable as a matter of prudence to give such a direction whenever there is a conflict between the prosecution and accused’s evidence (Salmon v R  WASCA 270; R v Chen, Siregar & Isman (2002) 130 A Crim R 300).
77. In such circumstances, it may be of assistance to tell the jury that:
- If they believe the evidence of the accused, they must acquit;
- If they have difficulty in accepting the evidence of the accused, but think that it might be true, they must acquit — because they will have a reasonable doubt about the prosecution’s case; and
- If they do not believe the accused, they should put his or her testimony to one side, and determine, upon the basis of the evidence they do accept, whether the prosecution has proved the accused’s guilt beyond reasonable doubt (R v RP Anderson  NSWCCA 488).