Volvo in another safety demo blunder

[Jason[98.EL]]

taken from carpoint

see the below link for video

http://ninemsn.carpoint.com.au/news/...-blunder-21809

Originally Posted by carpoint

Volvo in another safety demo blunder words - Joshua Dowling
New pedestrian warning system fails during media demonstration
The accuracy of Volvo's world first pedestrian warning system has been questioned after it mistook an inflatable human dummy for a tractor.

It is the Swedish car maker's second embarrassing failure during a media demonstration of crash avoidance technology in five months; in May a Volvo crashed into the back of a truck instead of stopping short of the impact.

The latest incident also uncovered the fact that Volvo engineers discreetly modified the inflatable dummy to 'trick' the car's radar system into detecting a pedestrian.

The latest bungle happened in the car park of a wine-making facility near Verona in Italy last Friday (September 24, 2010), during a demonstration of Volvo's new pedestrian warning technology at the international media launch of the new V60 wagon.

The system is designed to stop short of a pedestrian if the driver doesn't apply the brakes in time.

However, after more than 650 demonstrations during the launch over the previous four weeks, the dummy was knocked over three times in front of the very last media group -- three journalists from Australia -- who caught each mishap on video.

There was no damage to the car or the dummy in the latest publicity stunt gone wrong -- but Volvo's reputation for safety demonstrations took a hit.

"I am very sad this has happened," said Volvo's head of active safety Jonas Tisell. "We know the technology works, but I am not happy with the demonstration methods we are using."

An autopsy revealed that Volvo had discreetly inserted a small plastic panel covered in aluminium foil inside the sand bag near the inflatable dummy's feet to help the car detect the 'pedestrian'.

However, the dummy blew over in the wind before the initial doomed run and the discreet radar reflector was not put back in place correctly. Volvo also suspects a tractor and trailer directly behind the dummy -- but some distance away -- may have confused the pedestrian warning system.

However, the videos show that the tractor was only in the background during one of the three doomed runs. The video of the first run (taken from inside the car) which had the tractor in the background was yet to be uploaded onto the web when this article was published.

"The radar goes straight through the air-filled plastic dummy so we put some aluminium inside in the sandbag at the dummy's feet so the radar could detect it," said Volvo's active safety expert Jonas Tisell.

"When the dummy was knocked over, the reflector was not facing exactly the right way and instead the system detected the tractor directly in line of sight in the background.

"The camera told the system that there might be a pedestrian there because it detected the shape of a human, but when the radar went to search for it, it only picked up a faint signal for the dummy and got a more solid signal for the tractor."

Despite fiddling with the dummy and its stand, Tisell said Volvo had not 'cheated' the test.

"The problem with an inflatable dummy is that it's plastic and has air inside. The radar needs an echo, it looks for the body mass, and the radar doesn't bounce off an inflatable dummy. It cannot see it. It sees through it. So we added a radar reflector -- a 10cm piece of plastic covered in aluminium foil -- in the sand bag at the feet of the dummy.

"The system works on humans. With humans you have a mass the radar can detect."

Tisell said he has personally taken the place of the test dummy during development of the system.

"I was one of the volunteers to have the car drive towards me during development," he said.

But Volvo has so far refused to allow media to stand in front of the car during a demonstration.

"You never know what can happen. Driving towards a human being, if something goes wrong, it will be a very difficult situation to handle," he said.

"This system does not always work. It is a support system. It is not intended to replace the concentration of the driver.

"It is 90 per cent accurate. If we made it 100 per cent accurate it would stop for things it shouldn't stop for; a plastic bag blowing across the road. It should only brake when you have a real threat."

Tisell insisted there was no need to make any changes to the calibration of the pedestrian warning technology, but Volvo will review its public demonstration procedures for its crash avoidance systems.

The 5000 dummies identical to the one at the centre of the latest bungle -- distributed to Volvo dealers worldwide -- will not be recalled. But they may get a new accessory: an aluminium belt that will help the system detect the dummy.

The Volvo S60 sedan and V70 wagon are the first vehicles in the world to be able to come to a full emergency stop from highway speeds -- and detect pedestrians while the car is travelling at up to 80km/h -- and avoid impact at speeds up to 35km/h.

A camera in the windscreen detects the shapes and movements of humans up to 55 metres in front of the car, while the radar unit in the grille scans the road ahead to calculate the distance between the car and the pedestrian -- before deciding whether or not to intervene.

Volvo says it has been working on crash avoidance technology since 1999. The first pedestrian safety prototype was tested in 2004 and finalised in 2007. Volvo spent the last three years testing the system over hundreds of thousands of kilometres around the world.

"We have tested this technology on millions of people, wearing all types of clothes, in all types of environments, on all types of roads," he said.

"We will not stop doing demonstrations of this technology. We are proud of it and we want to display it. But we need to review our methods."

HOW IT HAPPENED

After 650 international press had been rotated through the Volvo V60 wagon media launch over the previous four weeks, the final group of three journalists -- from Australia -- had their turn to sample the pedestrian warning technology.

Just before their first run in the car park of a nearby wine-making facility, the dummy blew over and was put back on his feet.

In the meantime, a tractor had stopped some distance away behind the dummy, but directly in the line of sight of the car.

The car was then driven at approximately 30km/h by a Volvo instructor towards the inflatable dummy -- but the car didn't apply enough brake pressure and knocked the dummy over.

After a second, successful demonstration by the instructor, journalist Joshua Dowling got behind the wheel and drove towards the dummy at 35km/h (the maximum speed at which the system can avoid impact), but the car again struck the inflatable dummy after the system did not apply enough brake pressure.

The inflatable dummy was put back on its feet and two successful runs were completed. Steve Colquhoun, from the Fairfax Media group, then made three clean runs.

A Volvo safety expert later said it was unusual the system worked with such a faint signal from the pedestrian dummy.

Then the editor of Top Gear Australia magazine, Stephen Corby, also made a clean run. But on his second attempt (while driving at 30km/h) he knocked the inflatable dummy for six because the brakes did not activate at all and the system failed to detect the air-filled 'pedestrian'.

After that, the demonstration ended. You can see the video footage at the links below.

How the system is supposed to work (tested at 30km/h).

[Danny]

I can totally understand and accept that it was having trouble registering a blow up doll, it makes sense that the system is used to reflecting mass, not air filled balloon type objects.

As the Volvo guy said, Its a backup and support device for the driver, not a be all and end all, or substitute for driver alertness.

The way I see it is if it works just once in real life and saves a son, daughter, mother or father, then its good enough for me, and certainly better than not having the system at all..

[mac_man_luke]

Their explanation makes sense...maybe they should use a better human analogue eg crash test dummy or similar

[FocusLX06]

That's it. If it saves one life (e.g. All the successful passes) it's worth the money.

[DJM83]

Only human after all eh. Better now then later

[Full Noise]

Originally Posted by Danny

I can totally understand and accept that it was having trouble registering a blow up doll, it makes sense that the system is used to reflecting mass, not air filled balloon type objects.

I know a few humans who I’d love to shove in front of it, just in case it didn’t work.

[Fled74]

How about a pig carcass, like the Mythbusters use?

[UU TS50]

At least the dummy landed on its feet

[xtremerus]

They use a blowup doll so it won't damage the car when the system has trouble detecting a blowup doll
But will work with mass in the dummy. Another
Use a solid doll, Volvo. Or don't you have faith in the system

[GasoLane]

Originally Posted by mac_man_luke

Their explanation makes sense...maybe they should use a better human analogue eg crash test dummy or similar

Vote 1 for Harold Scruby.

After all it's his pedestrians they're trying to help so who better to help test it

[irish2]

More gimmicks to go wrong with the car. Will work great when it automatically brakes for a cat and you get cleaned up by the truck behind you.

[MAD]

Originally Posted by GasOLane

Vote 1 for Harold Scruby.

After all it's his pedestrians they're trying to help so who better to help test it

Exactly my first thought too!!