Software
Feasibility study
After the idea, I wondered whether I had knowledge necessary to developp a prototype in a correct delay, considering that I've a full-time work to maintain. All that from a "product" point of view, meaning that maybe one day I could commercialize it.
In software side I knew that I was able to do it, but hardware side was less trivial to me so it was necessary to study the question more deeply. Thus after several tests using I/O cards of the market, I realized that the most difficult would be the calibration, but not the development.
At this point, the realization of a prototype seemed to be possible.
After the idea, I wondered whether I had knowledge necessary to developp a prototype in a correct delay, considering that I've a full-time work to maintain. All that from a "product" point of view, meaning that maybe one day I could commercialize it.
In software side I knew that I was able to do it, but hardware side was less trivial to me so it was necessary to study the question more deeply. Thus after several tests using I/O cards of the market, I realized that the most difficult would be the calibration, but not the development.
At this point, the realization of a prototype seemed to be possible.
Software development
Following this feasibility study, I started out during the summer 2006 to code the main kernel of principal data computing and display. I drew an input simulator as well as a basic computing & display engine, but arlready covering the minimal needs of a dashboard : skinability and inputs dependency allowing to install it into other cars.
The software has been made under Windows XP, using OpenGL only. The objective of this is to be able to make it run in Linux if it becomes a market product, because the Linux licences are less expensive than the Microsoft ones.
And here is a screenshot of what I obtained :
As you can see, the look seem a little older. In fact it is about a copy of the standard look used of the dashboards of Renault Super 5 cars. The reason in is very simple, it acts of the car where the first prototype will be tested.
Following this feasibility study, I started out during the summer 2006 to code the main kernel of principal data computing and display. I drew an input simulator as well as a basic computing & display engine, but arlready covering the minimal needs of a dashboard : skinability and inputs dependency allowing to install it into other cars.
The software has been made under Windows XP, using OpenGL only. The objective of this is to be able to make it run in Linux if it becomes a market product, because the Linux licences are less expensive than the Microsoft ones.
And here is a screenshot of what I obtained :
As you can see, the look seem a little older. In fact it is about a copy of the standard look used of the dashboards of Renault Super 5 cars. The reason in is very simple, it acts of the car where the first prototype will be tested.
Why use such and old car as platform of test & development ?
1. It is my car, I like it and even if I had the means to buy some more recent, I prefer to keep it because I know that I can repair it (I like mechanics and to repary my car by myself). Moreover it is robust, highly reliable and nobody wants to steal it to me. Lastly, it does not draw the attention and the policemen never stops me !
2. The challenge was greather by trying to get & compute input data from analogical sensors or to even make them by myself. Opposite to the modern cars where these data is just served already computed (through CAN bus).
3. Personal, but that engages only me, I like the look "retro-high-tech" : the materal of James Bond 2028 into an obsolete car creates a contrast "high tech camouflage", which I particularly like.
4. Lastley, that did not form part of my development budget to acquire a new car...
1. It is my car, I like it and even if I had the means to buy some more recent, I prefer to keep it because I know that I can repair it (I like mechanics and to repary my car by myself). Moreover it is robust, highly reliable and nobody wants to steal it to me. Lastly, it does not draw the attention and the policemen never stops me !
2. The challenge was greather by trying to get & compute input data from analogical sensors or to even make them by myself. Opposite to the modern cars where these data is just served already computed (through CAN bus).
3. Personal, but that engages only me, I like the look "retro-high-tech" : the materal of James Bond 2028 into an obsolete car creates a contrast "high tech camouflage", which I particularly like.
4. Lastley, that did not form part of my development budget to acquire a new car...
Why Windows XP ?
The only reason is that I'll be able to develop quickly : I know it better and the drivers of the I/O board which I had tested are unfortunately not available or supported for Linux.
On another hand, the development was carried out in the optics of being able to carry it towards Linux or any other OS more real-time specific, if the needs appears.
The only reason is that I'll be able to develop quickly : I know it better and the drivers of the I/O board which I had tested are unfortunately not available or supported for Linux.
On another hand, the development was carried out in the optics of being able to carry it towards Linux or any other OS more real-time specific, if the needs appears.
Other skins ?
Yes, it is completely feasible, but for the moment I devote myself to the basic functionalities which I want to provide to this software. Even if that would not take much time, my time is invaluable and the developments are long. But it will come...
Yes, it is completely feasible, but for the moment I devote myself to the basic functionalities which I want to provide to this software. Even if that would not take much time, my time is invaluable and the developments are long. But it will come...
To adapt it to a CAN bus (ODB-II) ?
Who can more can less : to do that the only thing to do is to replace the computing module by a CAN input one, without modifying the rest of the software. If it is not currently done but it is only a question of time and means to test it.
Who can more can less : to do that the only thing to do is to replace the computing module by a CAN input one, without modifying the rest of the software. If it is not currently done but it is only a question of time and means to test it.
Summary of the characteristics
- Sensors computing module (raw, analysis, filtering, smoothing, etc...);
- Display module allowing to display radial gauges, animated indicators, counters, and display text;
- Skin management module.
Language : C++
Graphics : OpenGL
Operating System : Windows XP
Structure : Real time embedded : no dynamic allocation (easly adaptable to a static architecture), no threads, Doxygen documented code, etc...
Performances : 70-100Hz (frequency of data refresh and screen refresh, in other words, the duration of a complete cycle is 10-15ms).
- Sensors computing module (raw, analysis, filtering, smoothing, etc...);
- Display module allowing to display radial gauges, animated indicators, counters, and display text;
- Skin management module.
Language : C++
Graphics : OpenGL
Operating System : Windows XP
Structure : Real time embedded : no dynamic allocation (easly adaptable to a static architecture), no threads, Doxygen documented code, etc...
Performances : 70-100Hz (frequency of data refresh and screen refresh, in other words, the duration of a complete cycle is 10-15ms).
History
Summer 2006
I devote my summer holidays to develop a skinnable software allowing to compute and display the esential data in a dashboard.
October 2005
The idea was born. While I rolled on a right and tedious road.
Summer 2006
I devote my summer holidays to develop a skinnable software allowing to compute and display the esential data in a dashboard.
October 2005
The idea was born. While I rolled on a right and tedious road.