Jaguar I-Pace Forum banner
1 - 12 of 12 Posts

·
Registered
Joined
·
6 Posts
Discussion Starter · #1 ·
I use advance cruise control a lot. I think that the cruise control can be improved with the following:
- Ability to set the max speed when car stops. You can now only set the speed once you drive >20 km/h.
- Once you use the brakes you can only resume if you exceed >20 km/h. Why not resume if the car stops? This will enable me to let go of the brake paddle. Now I'm 'afraid' to use the brake when I use advanced cruise control, as I cannot resume if the speeds drops below 20 km/h.
I used to have a Volvo that had no problem with this. Am I the only one that sees this as an improvement point?
 

·
Registered
Joined
·
53 Posts
Darth_Fader said:
I use advance cruise control a lot. I think that the cruise control can be improved with the following:
- Ability to set the max speed when car stops. You can now only set the speed once you drive >20 km/h.
- Once you use the brakes you can only resume if you exceed >20 km/h. Why not resume if the car stops? This will enable me to let go of the brake paddle. Now I'm 'afraid' to use the brake when I use advanced cruise control, as I cannot resume if the speeds drops below 20 km/h.
I used to have a Volvo that had no problem with this. Am I the only one that sees this as an improvement point?
I am new to adaptive cruise control on my I-Pace (very used to non - adaptive) and I agree the improvement you suggest would be useful. It should be capable of being added with a software update?
 

·
Registered
Joined
·
84 Posts
It also needs to be less 'jerky' in steering corrections/more smooth. In addition, I've had several occasions with city driving that the car just did not brake and I had to intervene quite hard (have never had that happen on highways).
 

·
Registered
Joined
·
2,283 Posts
Getjaggywithit said:
It also needs to be less 'jerky' in steering corrections/more smooth. In addition, I've had several occasions with city driving that the car just did not brake and I had to intervene quite hard (have never had that happen on highways).
The vehicle in front has to have been previously detected as moving. It doesn't work if the vehicle was stopped before it comes into range of the Jag's sensors.
 

·
Registered
Joined
·
84 Posts
Goshdarnit said:
Getjaggywithit said:
It also needs to be less 'jerky' in steering corrections/more smooth. In addition, I've had several occasions with city driving that the car just did not brake and I had to intervene quite hard (have never had that happen on highways).
The vehicle in front has to have been previously detected as moving. It doesn't work if the vehicle was stopped before it comes into range of the Jag's sensors.
That must be it! And also one big fat bug imo.
 

·
Registered
Joined
·
528 Posts
Getjaggywithit said:
Goshdarnit said:
Getjaggywithit said:
It also needs to be less 'jerky' in steering corrections/more smooth. In addition, I've had several occasions with city driving that the car just did not brake and I had to intervene quite hard (have never had that happen on highways).
The vehicle in front has to have been previously detected as moving. It doesn't work if the vehicle was stopped before it comes into range of the Jag's sensors.
That must be it! And also one big fat bug imo.
it's not a bug, it's a characteristic of all radars. The receiver gets signals not just from the car in front, but also signals bounced back from side angles ("side lobes") and the most important being the ground about 15" below the sensor head. That's a rather large signal and is a static/non-moving target. I don't know specifically how these radars are programmed, but I suspect that once a valid signal is found that the receiver processor "locks on" to that signal and tracks it using differential velocity & range (relative speed and distance).

When the car approaches a static car in front it has to discriminate that new signal from the large side-lobes from ground reflections. That's very difficult, effectively the radar cannot "see" the new static target in front so it has no discrete signal to respond to, and at that stage it is not tracking so has no information on relative speed and distance.

When it approaches a car that is moving, the signal from that car has differential velocity and differential range, so the processor can distinguish a new signal that is valid because it is not behaving like the noisy reflected ground signal.

I'm not sure my explanation 100% matches the how these radars work, but my BMW suffers exactly the same way and coming from a radar background Ive sort of worked out that the problem must be related to that zero-speed ground reflection. There may well be other manufacturers that use a different design of signal processor, and there are ways to improve this, but not by software alone. (The radar receiver signal processing is not all done by Jaguar car software, some of it is embedded in the hardware processing boards).
 

·
Registered
Joined
·
2,283 Posts
Yeah it is just a feature of pretty much every adaptive cruise control system. A radar has no way of working out whether a static signal is a car, rising ground, an overhead sign board, a traffic light gantry or whatever. So they have to filter them out or the car would be slamming the brakes on every few minutes. It's a weakness that all manufacturers put up with. The only one I know of that does it differently is Tesla who uses an array of cameras. They have their own issues but that system amounts to way more than adaptive cruise control.
 

·
Registered
Joined
·
84 Posts
Billy Bunter said:
Getjaggywithit said:
Goshdarnit said:
The vehicle in front has to have been previously detected as moving. It doesn't work if the vehicle was stopped before it comes into range of the Jag's sensors.
That must be it! And also one big fat bug imo.
it's not a bug, it's a characteristic of all radars. The receiver gets signals not just from the car in front, but also signals bounced back from side angles ("side lobes") and the most important being the ground about 15" below the sensor head. That's a rather large signal and is a static/non-moving target. I don't know specifically how these radars are programmed, but I suspect that once a valid signal is found that the receiver processor "locks on" to that signal and tracks it using differential velocity & range (relative speed and distance).

When the car approaches a static car in front it has to discriminate that new signal from the large side-lobes from ground reflections. That's very difficult, effectively the radar cannot "see" the new static target in front so it has no discrete signal to respond to, and at that stage it is not tracking so has no information on relative speed and distance.

When it approaches a car that is moving, the signal from that car has differential velocity and differential range, so the processor can distinguish a new signal that is valid because it is not behaving like the noisy reflected ground signal.

I'm not sure my explanation 100% matches the how these radars work, but my BMW suffers exactly the same way and coming from a radar background Ive sort of worked out that the problem must be related to that zero-speed ground reflection. There may well be other manufacturers that use a different design of signal processor, and there are ways to improve this, but not by software alone. (The radar receiver signal processing is not all done by Jaguar car software, some of it is embedded in the hardware processing boards).
Must admit I am not familiar with how it works or is supposed work, so I appreciate this information. It's the first time I have ACC and for as an "end-user" I would expect this to work with signals bouncing off (from what are pretty large) objects in front, stand-still or not. And Also thought the camera would be in on it judging by the brochure. It's no big deal, just expected otherwise.

Water Automotive tire Vehicle door Font Asphalt
 

Attachments

·
Registered
Joined
·
528 Posts
Getjaggywithit said:
Billy Bunter said:
Getjaggywithit said:
That must be it! And also one big fat bug imo.
it's not a bug, it's a characteristic of all radars. The receiver gets signals not just from the car in front, but also signals bounced back from side angles ("side lobes") and the most important being the ground about 15" below the sensor head. That's a rather large signal and is a static/non-moving target. I don't know specifically how these radars are programmed, but I suspect that once a valid signal is found that the receiver processor "locks on" to that signal and tracks it using differential velocity & range (relative speed and distance).

When the car approaches a static car in front it has to discriminate that new signal from the large side-lobes from ground reflections. That's very difficult, effectively the radar cannot "see" the new static target in front so it has no discrete signal to respond to, and at that stage it is not tracking so has no information on relative speed and distance.

When it approaches a car that is moving, the signal from that car has differential velocity and differential range, so the processor can distinguish a new signal that is valid because it is not behaving like the noisy reflected ground signal.

I'm not sure my explanation 100% matches the how these radars work, but my BMW suffers exactly the same way and coming from a radar background Ive sort of worked out that the problem must be related to that zero-speed ground reflection. There may well be other manufacturers that use a different design of signal processor, and there are ways to improve this, but not by software alone. (The radar receiver signal processing is not all done by Jaguar car software, some of it is embedded in the hardware processing boards).
Must admit I am not familiar with how it works or is supposed work, so I appreciate this information. It's the first time I have ACC and for as an "end-user" I would expect this to work with signals bouncing off (from what are pretty large) objects in front, stand-still or not. And Also thought the camera would be in on it judging by the brochure. It's no big deal, just expected otherwise.

https___www_jaguar_com_Images_Jaguar-I-PACE-Brochure-1X5901910000BXXEN01P_tcm660-627215_pdf.png
interesting reference to the camera....... I suspect that's more for the lane control (via white line detection)

on the BMW there are 2 radar head patterns, and the ACC takes pre-processed signals and further makes the actual driving decisions (such as speed up, slow down etc). It also takes steering wheel angle into account, so looks slightly left when going around a left hand bend etc.

Problem with "signals bouncing off objects in front" is that radar is a 2 way system.. signal sent out is diluted by the square of distance, and the reflection is again diluted by square of the distance coming back. So there's a power of 4 reduction in signal. It means that something 10 meters away gives a radar signal that is 10,000 times less than something that is 1m away. Typically the ground reflection is about 30 cm from the sensor; a car in front becomes important when it is about 50-100 meters in front of you. The ground signal is 150-300 times closer, so the ground signal is about 100 million times stronger !! That's why it's so difficult to resolve decent static targets (as previous poster also comments about signs, kerbs, etc)
 

·
Registered
Joined
·
528 Posts
Tophe74 said:
They could use front camera to help radar
they could, I think that's what Tesla do in Autopilot cars.

It would take a whole lot of new software to do the "car recognition" and would start to put these cars into Model S price category........
 
1 - 12 of 12 Posts
Top