10.0.3 feedback and suggestions

The maps get blank many times i think that the issue is not solved yet and the indicators too it turns off too early before I take a turn. When I apply both brakes to take a turn and the indicator switches off too early before I take a turn please solve these 2 issues at next update it is happening many times

Yes I also have faced this problem the indicator switches off before taking a turn or just if I apply brakes it will turn off I need to turn it on again​:pensive::pensive:

As @blackworks said the indicator feature works based on a sudden incline or impact that sensor gets. For me the auto cutoff works most of the times 80%. It depends on riding.

But as it is not expected that everyone rides the same way, it is recommended and compulsory that they give an option to weather user want it or not. Forcing to use it is not required.

Let the machine learning and improved algorithms take some time and improve the correct auto indicator turnoff feature. Because the percentage users facing issue is high that those who are not facing(these are very less may be negligible when compared).

Also option to select the sound level is also required or keep it at previous levels only. Obviously safety is important than comfort.

1 Like

I believe this would have to do more with the actual IMU calibration rather than any algorithm. The algorithm is only going to use the IMU output to make a decision. If the IMU output isn’t consistent, then the algorithm is going to fail.

One more bug @abhishek.balaji

Are you referring to the lack of satellite imagery in the map background? If yes, it’s most likely because of poor network coverage and the bike wasn’t able to download that image from Google maps.

I fail to see the point in wasting resources trying to teach a machine to do what humans are wired to do subconsciously. As far as solutions to nonexistent problems go, this auto cutoff is the equivalent of a pre cracked egg.

images

Brings us back to the original question: is there a point and purpose to this dumpster fire?

2 Likes

Buts its my daily route it loads everyday suddenly all of a sudden it fails

These are not like preloaded maps stored in vehicle . When ever we give a destination it fetches data from google map API and give you the route and the route gets updated on dashboard as we go forward on the route. As @raghav.srinivasan said this is due to the network picked up by the vehicle. So it has nothing to do with if you are going on a regular route or new route.

This also happens if we are crossing any defence related area as suddenly your network reachability reduces. If we restart maps it loads or works slowly though as we stop aside and refresh it, but while on the go it cannot pick the proper signal in such areas as per my experience in this matter (I faced this and above statements are based on my observation & understanding of the matter).

1 Like

Maps can be cached on frequently travelled routes, assuming there is enough storage available.

2 Likes

Yes that is true but my point is that here it is not happening like that. We are able to mark favourite places but not favourite routes. I might also be wrong but that is my understanding.

1 Like

I might differ on this because

  1. If IMU calibration is the issue then this would have been sorted out (or may be improved) way long back in new vehicles at least. Also in old vehicles they could have sorted it out while service.

  2. If it is not an algorithm related there is no point is saying we are working on it for next OTA.

  3. Sensors or units designed by putting a lot of engineering in it doesn’t fail in general it it does then it is an engineering failure even if it is a calibration related it is engineering failure.

  4. Sensors do what they do but it is upto the algorithm that tells the plc or the device intelligence which signal from the sensor to be captured. It is all about capturing a right combination in the algorithm to tell the intelligence that yes if this happens turnoff my indicator.

Just take an example of self driving cars in initial days they do crashed many times but that doesn’t mean that the problem is with sensors or something it is about improving the algorithms over a period of time they improved the system.

A lot of things can be done in theory, but aren’t done in this case, probably because of an underspecced display unit. This is basically a smartphone embedded in the vehicle. At this point, an actual smartphone tacked over the display would be a more viable option.

IMUs are known to be quite noisy and it apparently takes a lot of software filtering to clean up the noise. That said, auto canceling indicators in cars are easier to implement as opposed to bikes because of one simple reason: range. In a car, the steering has almost 900 degrees of rotational range for a relatively large turning radius. This gives you a lot of resolution to figure out the driver’s intent when you detect a change of steering angle and fairly accurately predict the completion of a turn. In a bike, you have maybe 20 degrees of angular range to figure out the same thing. Good luck with that.

1 Like

This is what I meant that it has more to do with the software and less on the hardware side. As hardware does what it does. It is on narrowing out on what angle and range to be taken as criteria to cutoff and may not be the issue of calibrating IMU.

This very apt. No point comparing a 4W auto cancellation system with a 2W because your handle bar does not travel far enough for the vehicle to trigger a contact switch.

Coupled with the fact that a lot 2w riders don’t even bother to stay in the same lane while turning, the IMU isn’t going to find the lean angle sufficient enough to trigger a shut-off. Tweaking the threshold would most likely trigger false positives because now even a simple change from one end of the lane to the other gets detected by the IMU, initiating a shutoff signal.

With my 450, touch wood, the auto cancel works most of the times!. I stay within the lane and tend to lean slightly more than I did on the Activa (more confidence with the lower CoG).

Absolutely bang on. Unethical first, professionalism be damned.

But it worked decently before the latest update so it can’t be all the hardware’s fault.

Exactly … feels more like they’ve alternates the settings. I almost never switch to Sports mode… rather stick to ride and if I need some more power, go to Warp.

3 Likes

Hey folks,

Update on the auto indicator issue: Internal updates - Ather 450X & 450 Plus

On the sport mode vs ride mode discussion here, we’re reaching out to some of you to pick up the vehicle and check for this issue back in our office. We’ll keep you posted on the findings and any resolution.

4 Likes

Thanks a lot @abhishek.balaji @ather for listening to us. :pray::pray:

I appreciate both the moves. :+1:

2 Likes

My Sincere Request on Indicator Sound. Have full sound or no Sound option. Current Sound level is Totally Useless! Please takes this as priority!

2 Likes