Examples from the Oops! File

Some 'Oops' are as simple as the wrong sign giving the wrong message. Zebra crossing signs on a signalised left turn slip lane?

The examples below are a little more serious . . .

Here are some examples (A to F) where things have gone - or continue to go - badly wrong. Things can go wrong (meaning they create a road safety problem) at an individual site, in a large area or with a 'new' concept, the safety implications of which are not adequately understood.

If any of these issues sound familiar and you'd like to discuss your issues with me, do get in touch.


A roundabout was about to open. The Council was concerned that drivers could effectively drive straight through at speed, down hill through the roundabout. I undertook a road safety audit. It recommended, among other things, that the layout be physically tightened on the down hill approach, so drivers will be going slow enough to give way if required. My recommendations were then implemented. The roundabout has operated successfully since.


In 2009 Dumbaugh & Rae remarked in 'Safe Urban Form' that "While often romanticized by contemporary proponents of neotraditional development, the urban grid was popular in the 19th century not to encourage pedestrianism, but to promote rapid land development."

The re-introduction of local street crossroads (whether give way controlled or uncontrolled), as part of re-introducing gridiron form of urban development, is a negative shift for road safety. One self-proclaimed 'expert' has said there is no safety problem with local crossroads. But as the above table by Elvik (2009) shows, the crash rates for crossroads are generally double that of T-junctions and at local street intersections are up to three times greater - or six times greater than where roundabouts are used. The ongoing introduction of new crossroads into the road and street network is a ticking time bomb.


A Council scheme to beautify a local shopping street included the treatment on the left: a pair of high wall with paved brickwork across the road immediately after. The wall hid the pedestrians from the drivers' view and the brick paving gave pedestrians the idea they had priority, which was not supported in law. After I advised the Council of the hazards, the old wall was removed and rebuilt much lower (on right). The paving was replaced by a conventional zebra crossing that both drivers and pedestrians understand.

A road safety audit of the original design by an experienced road safety auditor would have saved the Council money and saved the community trauma.


An existing intersection had to be signalised, due to increased development in the area. The intersection design (see image by Nearmap, below) was carried out by a large traffic engineering firm (seriously) and then the design was audited by a team of reputable road safety auditors. They found an issue for cyclists, but advised that no major safety issues were identified from the perspective of motorists. They did not recommend any changes except for cyclists (seriously). The new design was built. Just before it opened a smaller company audited the project. They identified one line marking problem, but proposed one sign to solve it (seriously). Then the intersection fully opened. That's when the Council started receiving phone calls from irate drivers . . .

Drivers doing the double left turn were conflicting with each other (side swipes). Drivers doing the right turn in the opposite direction were cutting across and conflicting with those left turners. Drivers were travelling straight through in the right turn lane (shown at the lower left here), conflicting with drivers from the left lane when they had to merge.

In response, the Council called me in to carry out a road safety audit. Had I been engaged to conduct the original design audit, all the following problems could have been identified and avoided:

  • The strange chevron markings on one corner (middle of this image) are pointless. They are not needed.
  • The turning lines are too tight (because of the chevron markings) and cause drivers to stray out of their lane.
  • There is no standard line marking at the start of the right turn lane (out of view, to the right of this image) to keep through vehicles out of the right turn lane.
  • The crosswalk over the side road is too far away from the corner (risk to pedestrians from left turners from the main road).
  • Travelling to the right (in this image) there is no warning/advice on the need to form one lane. The merge taper is too short.
  • There is a bicycle box marking for right turns from the side road: how can that operate safely or legally?
  • At the 'Keep Clear', right turns can be made into the side street. This risks rear end collisions (The old intersection had a separate space for right turners to prop).
  • Out of sight to the left, the left-bound lane veers out from the kerb, around some redundant line marking, before veering left past a refuge island.

The two previous audits were conducted on behalf of developers. One of my recommendations to Council was that they need to establish policies and procedures that keep them more in control of these types of developer-funded projects.


A bridge deck needed to be replaced. The project simply kept to that objective, rather than being thought about in a broader traffic context. The approach road in both directions is four lanes wide - though, as is so common on many of Melbourne's undivided four-lane-wide arterial roads, it is only marked with a centre line. I was involved only after the construction contract was awarded: I had to audit the roadwork traffic management and conduct a post-completion road safety audit.

Even how the road safety audits were organised was the wrong concept. Policy required only one audit to be done, so it was decided by the road authority to do it at the end. That is the opposite of what the Austroads guide says: get a road safety audit done as early as possible in the design and development process.

Here is a photo of the bridge when I first saw it (during construction):

On the other side of the road, the kerbline follows straight through. On this side it has always necked in, to provide space for the footpath. Yet the substructure of the bridge is the same under the footpath as under the road: it could carry a widened road pavement - and a separate footbridge could be built beside it.

It's the wrong initial design concept. This is more amazing, because the stream under the bridge is the border of two local government areas, so three road authorities were involved. Yet no one thought to widen the bridge. An audit at the design stage would have identified the safety hazards with having one-and-three-quarter traffic lanes in each direction.

Below are photos taken during the post-completion audit:

With one-and-three-quarter traffic lanes each way, the risk of side swipe crashes - or worse - is ever present. Drivers find it difficult to identify how many traffic lanes there are in conditions like this. To overcome these crash risks, my post-construction audit report recommended that the road be marked into lanes: in one direction, two full-width marked lanes need to be marked throughout (coming down the hill, over the bridge and beyond). In the other direction there should be a standard lane drop/merge approaching the bridge (one lane on the bridge, two lanes marked elsewhere). With this, there is room on the bridge for a half-metre-wide painted median to separate oncoming traffic. This was recommended.

To the best of my knowledge these recommendations have not been acted on and the road still looks like it is in the photos above.

This is a ticking time bomb.


The biggest OOPS! of all. It's over a large area and it will last forever.

When the Armstrong Creek area, on the southern edge of Geelong, Victoria was about to be developed, a concept of having frequent low-key distributors (i.e. 'connectors') was developed, rather than providing fully-developed, separate arterial roads (without direct driveway access) at 1.6 km spacings. This is what got approved and is now being built.

Here's what is happening (above). The black and red lines are existing arterial roads and green is a freeway. There will be no new east-west or north-south arterial roads in the area. New road connections across the railway are not being provided. There will be two-lane 'connectors' with direct driveways to houses, at 800 m intervals.

Here (above) is what should have happened. Had VicRoads not earlier deleted a section of its Traffic Engineering Manual that advised this is exactly what is needed in new urban areas (see my Free Download on functional road hierarchies), it might have occurred. Instead, the unsafe town planning fad of low-key distributors prevailed.

This dangerous town planning nonsense is well described in a report by Brown Consulting in 2011:

‘The City of Greater Geelong adopted a road network strategy to interconnect arterial roads with more frequent, narrow, connector roads at 800 metre spacing throughout the Armstrong Creek Urban Growth Area. The connector [i.e. Distributor] roads are generally envisaged to comprise of one traffic lane in each direction, with generous footpaths, shared or dedicated cycle paths, and kerbside parking provided on both sides. These [connector roads] will distribute traffic evenly, avoiding almost entirely the need for major, heavily-trafficked roads which are typically unpleasant and unsafe for pedestrians and cyclists, and create a barrier between development on either side.’ and

‘All roads within [the area] are proposed to be connector or lower order roads …’

Such noble thoughts. Such ignorant nonsense.

In 1989 Dr Ray Brindle (town planner and engineer) at ARRB set down fourteen conditions by which any new development would be deemed to satisfy traffic performance, amenity and road safety requirements. Conditions 3 is:

'There are no direct frontages to arterial or district distributor roads, or to connectors.’

In 1986 the Federal Office of Road Safety, in its Safety of Older Pedestrians report said:

'Older pedestrians are most at risk on ‘sub-arterials’ [a.k.a. collectors or connectors], rather than on the busier main roads where it appears that measures to assist traffic to interact safely also benefit pedestrians.'

The Armstrong Creek area will now be blighted forever with higher rates of road crashes and daily congestion. The existing arterial roads will also be more heavily congested than if separate arterial roads at 1.6 km spacings had been provided and connected over the railway. It will be more like a slow, lingering death than a ticking time bomb.

Who benefits from this sort of misguided town planning? It is difficult not to conclude that the only significant beneficiaries are the companies that develop the land. With no separate arterial roads (without driveways) having to be provided, there is more land left to subdivide and sell.

It is quite evident that road safety was not considered in any meaningful way when Armstrong Creek was being developed. Road safety engineering is a separate discipline from other engineering and it needs to be involved just as much as other separate disciplines. There are simple processes that can be used to involve road safety engineers in the early town planning considerations. These are discussed in the road safety audit training workshops that I conduct.

To return to the Traffic Engineering page Click here. To go to the Contact me page Click here