RaceChrono Wish List

123468

Comments

  • Hi

    I would like if Race Chrono will look in the future something like this:(Windows version)
    http://www.bros.co.kr/v3/products/digital.php

    Your team is on the right way to do this.
    Thanks.
  • Very nice find, thank you! Lot of fresh ideas there!
  • edited November 2009
    Hi aol!

    As I seen that you wrote earlier about plans to add accelerometer function I want to show you this small symbian (sisx) program that uses the acceleromter in a nokia 5800 (i guess most new phones has these sensors used for tilting the screen theses days?! Most new nokias the last year or two have of the upper/mid segment...) Check out one program at http://www.creabros.fi/cms/index.php?id=28 and source code and a simple numerate version of the same at http://www.creabros.fi/cms/index.php?id=27 ( the nokia sensor api link is here: http://wiki.forum.nokia.com/index.php/Nokia_Sensor_APIs)

    The program is able to use the tilt and also acceleration movments of the phone to move a small ball on the screen and it works suprisingly well, the second shows the sensor values just in raw values. The sensor is not what I can see just working as a gravity sensor, but actually working as a good accelermoter sensor should, with this small program I think you can judge the quality of it. If you dont have a 5th ed phone I can film this and send to you if you want...

    My idea here is that pherhaps this internal g sensor is not the very best when it comes to be accurate in absolute values and it would maybe be hard to the calculate absolute real G values, but for sure it would be great benefits and work for obtaining the direction and exact actual acclerating directions with this sensor. My biggest intrest/idea for this would be to get much more accurate acceleration and braking points than you get with just the gps signal. As the sensor is very sensative it seams that it would be possible to simulate the throtle control and braking control to a much higher level than what i think the racechrono is possible today. I am a novice into racing but this would be great for me when studying corner and exit approches, like am I braking with constant push to the pedal, or braking by many smaller brakes inputs and thus loosing time, etc...

    As the source code is free pherhaps this would not be so hard to implement ( i have 3 nokia phones all with acc sensor inside 3rd ed fp1, 3rd fp3, and 5th ed so I would be willing to beta test and help to calibrate the application/sensor if that is needed,)

    Also many thanks for a great program!

    Regards Krste

    /Sweden
  • edited November 2009
    Hi Krstec!

    I have prototype version of RaceChrono that logs the acceleration data on 5th edition and 3rd fp2 phones.

    I've run in to problems with it and basically started doing something else. The data logging was easy to do, but calibration is hard. Basically you need to rotate first the X/Y/Z axis so that it corresponds the position the phone is attached. I tried to make it so that phone can be in any position, but detecting the position in RaceChrono was just too hard nut to crack for me. I don't want to require users to attach the phone in one exact position either so I need to find some nice solution to this.

    This is much easier for "performance testing only" products where the required accelerations are one dimensional. But for RaceChrono the accelerations need to be 2D.

    Also there is a problem of synchronizing the GPS data to the acceleration data. I haven't found a suitable algorithm to do this reliably.

    -Antti
  • edited November 2009
    hi again Antti!

    :-) Should have guessed that you already tried this! Oki i thought with the possiblity to input like odb signals that timestamping would not be an issue but i understand.

    When it comes to calibrating, would it not be enough if you as a user chose what is forward left and rigth directly onto the screen with some kind of gauge/circle with arrows pointing outwards around a circle on the screen just after placing it each time if you want the function activated and letting z always be the direction of the gravity vector at standstill? As i said my thaught about the sensor is that it is for sure not so good, but if you set the resolution to be somewhat low in the calculations and interpolate over a short time for the vector components, pherhaps divide the horizontal plane into as ex 18 sectors and still get a good estimation. Doing this way the calibration would also not be needed to be so accurate by the user either as this would give for in this case 20 deg of error margain. Sorry if my ideas sounds not to smart or you already have been thinking of them. I am just myself now thinking how more advance/comercial systems are fixing this issue, i assueme the combo of a gyro and acceleromerer is the proper way to go but as a comprimise now there should be some way...

    And i assume you tried also to calibrate the x/forward direction against the gps signal when the vhehicle is going in a straight forward line according to the gps? Or this would be very complicated to maybe...

    Krste
  • edited November 2009
    Hi Krstec, thanks for good ideas! Do you have any pictures/screenshots/etc of products in this kind of calibration mode. The arrow is not bad idea at all.

    Calibrating against GPS was something I tried but it was not reliable. It probably could be made reliable, but the amount of time I have for each feature is just not enough to do anything so elaborate.
  • edited November 2009
    Hi again

    I realise what you meant now that it is very hard to implement an easy and straight forward calibration method independent of positioning of the phone.

    In the end I guess that the easiest/quickest option to implement with a little effort for you (as you say you have some algorithms already written) would be that there should be possible to input the rotation of the phone manually in some menu as exactly as the user wants to do it, so maybe this manual input is enough. The problem with the gps synchronization, is it to at all align the accelerometer data, or the problem is to match g values to align with g values given by the gps data? As I said last time, I see the use for the performance user, that would be able to study throttle and brake control and maybe there is no need for a truly G sensor in an intial stage. Thus also it would probably be enought with having limited amount of directions 10-20 that the program is interpolating within.

    This looks like an interesting externa blue tooth accelerometer however but is little bit expensive 200$. http://www.sparkfun.com/commerce/product_info.php?products_id=8563
    I see that it has an inclosed gyro sensor also so that could give the yaw angle etc., also this is the third revision of the sensor, with documented specs and been used in many applications before. (I read before that for the obd you wanted to write code for reliable products that for sure would stay out at the market for a long time and with suport by the manufacturer.)

    I have looked into the pattern idea a little bit to during the day but I guess it is not really that easy either...

    I cant find an application that uses a system like I thought of. I just made an example pattern image and put it on the phone (check it at http://picasaweb.google.com/krstec/Arrows#) .

    My idea is as long as you have a rotation of the phone along its x axis, you will be able to tell what direction is pointing forward with a pattern like this whatever rotation you have along the y and z axis. So you put the unit in the car as you want, then you look at the phone so you are looking at it and also straight out of the wind shield along the X AXIS OF THE CAR (as I have positioned the phone on the pictures). Then you click at the arrow that you decide is pointing forward.

    Now knowing which direction is forward , together with the data from the accelerometer sensor which will give direction of the ground it is possible to calculate what is front, left and right of the car. However an issue arises if the rotation of the phone is zero along the y axis and rotated along the z axis for example, in this case however I would think that it would be easy for the user to put in the actual rotation of the phone along the z manually by just inputting the rotation angle manually. Then it is possible "for the software to calculate" what direction is pointing forward of the phone and thus you have your calibration.


    I wish I was more into programming so I could help out a little. But anyway,the program is so far great as it is now also!! regards Krste
  • I'm not sure if that arrow idea would work I will try to think about it though.

    External accelerometers would be much better as you can attach the unit semi-permanently on the car or bike. That sparkfun link looks mighty good. If it had 5 Hz GPS integrated in same unit it would be perfect :)
  • :-) yes it sure would! Starting to understand why even the cheap comercial systems are not that very cheap now hehe with all parts and bits needed to include...
  • edited November 2009
    Some time ago I suggest this, but now I found an image that can speak by self.

    http://www.devtoaster.com/products/rev/images/screenshots/shot3.jpg

    Will be usefull for rally stages, so you can see in the phone what type corner begin.

    Also will be nice for our favorite tracks to use backgrounds

    So the suggest is the possibility to use backgrounds autoaligned (or not) by GPS data exported from google maps.
  • edited December 2009
    I apologize if this have been mentioned before (or if I just can't find a suitable function/setting) but I feel it would be really useful in the teaching setting (driving schools etc).

    I would like to be able to compare Trap speeds and sector times of several laps at the same time, not just one lap at the time as today.
    The reasoning is that it would be valuable to show the difference in seed/time in relation to the route taken, and this directly on the Mobile screen. I realize that all laps cannot be shown but maybe browse through the laps in comparison with the chosen comparison lap.
    The important bit is to be able to show the effect of a different route than the optimal.

    But what an app, I'm really glad I don't bought a *box (no analysis without a laptop). :D
  • edited December 2009
    Hi Niklas! Do you mean you'd like the trap speed and sector time to be visible in the Live Timer? Instead of the split time only that is showing currently?
  • edited December 2009
    Hi aol,

    Not in the live time, I'm thinking about when you review the laps/sectors to also see the comparison lap in the route/map and not only in the speed graph (where the comparison lap is shown in dark yellow).
    This way it would be easier to see that the faster sector time can be due to a later apex or later turn-in.

    Having the sectortimes rolling in the Live Timer as you pass into the different sectors (full lap in different color) could probably help those who can look at the screen more than once per lap (placed mid view), but that is not me.
  • What I'd find useful is the ability to store setup data with the log. At the moment, RaceChrono allows you to store the driver name and a free-form notes field. What'd be great is to be able to add real data like spring rates, ride heights, tyre pressures and the like. Of course, moving on from there, it'd be great to be able to query the data to see where times improve and what's quicker as a result of various changes.

    This can all be done after the session but being able to change the data in RaceChrono would allow you to more easily compare where you're losing/gaining time, not just the fact that you are. Using lots of splits, it's currently possible to output the data as a CSV and import to a database with the setup data.
  • edited January 2010
    Nicklas: added to the wish list.

    Andy: Sounds quite complicated, as everybody have different setup needs. But if the user interface issues can be overcome it is quite nice idea... sort of data mining. Maybe it would be easiest add to notes what was changed in the car, and then just improve the RaceChrono analysis with a feature like "where am I faster with this set of laps, compared to the other set of laps".
  • Yes, that sounds feasible. Personally, I'd be more likely to add 100 splits to the lap, export as CSV and import into the setup database. Then I can write a query to show where I'm faster/slower on certain laps, based on the data. Perhaps it'd be useful to be able to colour the track map to show where one lap is faster than the other for comparison? Green = new lap faster; Red = old lap faster, with a graduated scale between the two. Not sure how easy that'd be though!
  • edited January 2010
    It's doable I think. A bit slow to calculate, but I think it's still feasible.

    Added new item to the wish list: "New display mode for laps. Show where you are faster than the comparison lap in green, and where slower as red."
  • Wish Racechrono on my Nokia N900 with Maemo ...
    BR Andreas
  • Andreas, N900 port will be relatively easy (unlike iPhone or Android which are hard), so it is coming.
  • How about a downloadable manual ? Perhaps pdf. Be nice to have this on the phone, on the move as you learn the app.
  • Full screen display, which only displays the seconds-part of the (1:33,4) laptime:"33,4".
    Background and font color user selectable. Maybe also font style and thickness selectable.
  • edited February 2010
    Hello to all,

    Add trap after the 1.35 beta : missing time, speed, altitude.

    In 1.35 beta when moving on "Traveled route" i see number of rec, time, speed, altitude.
    Depending on the position on the track and the data displayed (time, speed, altitude) can place the new trap.

    In versions after the 1.35 beta, you removed “add trap here” in route view (you add traps in “Browse traps” instead).
    I do not see these valuable data ... is less ! Why ?

    I think it would be nice to expand the possibilities of analysis to position the traps. For example colorize the trace functions accelerations positive or negative as in the "Hi/Low Speed" mode.
  • I removed it from there because I wanted to centralize all trap related action under one view (browse traps). I can add some more info there when I have time.
  • I understand because you have a mountain of things to do ...
    Do not get lost and concentrate on the essentials ... over time you can spend.
    But let me tell you RaceChrono is a very good analytical tool that can take anywhere.
    Thank you !
  • edited February 2010
    Did I mention before an idea to export track maps to EPS? It'd be great to be able to output the routes for use in other software or for printing track maps (to paper or sticker graphics).



    It's a plain-text file format that's well documented.
  • edited February 2010
    how about send SMS with latest times to a specified number, so say every 10 laps (or whatever) it send a text with the last 10 laps, whatever fits under 160chrs...something like "1:1.03.59,2:1.04.55,3:1.06:44...."

    A simple version of viewing laptimes on a remote device.

    we have been using recently for kart enduranace (3+hours) and its handy to know when a driver is getting tired....SMS would be perfect for that.
  • edited February 2010
    also....if doing a feature to show on another device remotely, then it should have the option to show multiple devices and compare, so it could be used as a complete timing solution with multiple cars, so for example 5 karts in a race with racechrono device on board, could replace the transponder system and show all times together, qualifying, grid position, laps completed etc, just start a "remote session" and join all the devices to that sesssion and look at a remote device will the the data on it?
  • Autopush session data to paired laptop.
    When speed is under 30kmh(user selectable); racechrono autopauses.
    And starts to sniff if laptop is in range.
    If laptop available, pushes the data to laptop.
  • edited February 2010
    If Racechrono would be able to transmit some useful data to another device, that would take it to another level above competition. We have GPRS , 3G or SMS on the phones already, might as well use it.
    If for example Racechrono transmits a small amount of data via GPRS once every lap and sends a Laptime, Max/Avg/Min Speed and possibly engine temperature to a server. Then a user can logon to a web page of the server and view the data "LIVE".
    Take a look at Nokia Sports Tracker for example http://sportstracker.nokia.com/nts/main/index.do
    The little SportsTracker app on the mobile phone transmits data to the server via 3G or GPRS so all the info can be viewed on the web page almost live. I have tested it and there is about 15 sec delay.
    MattNJ , that could be an interim solution for the enduro kart race, as it shows you laptimes on the Nokia web site. Obviously, you will need at least two mobile phones with Internet connection, one on the kart and the other one for team to view the data on the website.
    So, aol, how difficult to incorporate this sort of functionality into the Racechrono?
Sign In or Register to comment.