Friday, 4 October 2013

What we can (not) learn from F1: Driver-Car-Interfacing - Part1: Visualisation

Many teams get inspiration for parts of their car from F1 or other racing series. I am not saying that this is a bad thing in general, but whenever you are thinking of copying or carrying over ideas from somewhere else, you still have to question why it has been designed the way it is. These "What we can (not) learn from.." posts will critically and hopefully objectively assess things done in other engineering areas and evaluate whether it is good or bad to just take over the respective parts/ideas or whether FS needs a different approach.
PLEASE NOTE: These are my personal findings. Acting according to these posts does not gain you any points in the design event or guarantee a trouble free season/event etc. It is just build on my experience as former participant and current official.

This post is about Driver-Car-Interfacing. It is not about ergonomics, but about the driver exchanging information with the car and vice-versa. Back in the days there were mechanical gauges showing the current coolant temperature to the driver and the driver may have been able to open and close valves to change some kind of vehicle behaviour. Nowadays this usually goes electronically via displays, lights, buttons and switches. Therefore I will focus on dashboard and steering wheel design with respect to information transfer in this post.



Let us start with visualisation of information to the driver. What does the driver really need to see/be informed about? In a perfect world: nothing. All of the car's systems will always work perfectly well, none of the conditions will change, no parts will fail and the driver can easily hear when he (note, I will always assume a male driver in this post, which I do just for convenience reasons and not because I am a sexist) has to shift up or down. Wouldn't it be nice if it was that simple? Of course, in reality, this is not the case. Temperatures or pressures will be too high or too low, the driver will be inexperienced, thus not knowing when to shift, sensors will fail etc. Formula Student cars are prototypes and as such they need lots of maintenance and systems often fail. From that I conclude that the driver needs information provided by the car's systems. Again, in the perfect world, the track side engineer(s) will just evaluate the data received remotely from the car and tell the driver what to do, but this is often not the case for FS and even in professional motorsport this does not always work. Just remember the marshalling/telemetry debacle in F1 earlier this year: McLaren Electronics now causing trouble in Melbourne.

What does the driver need to see? Well, everything that would either make the car go faster or more realiable, both resulting in gaining more points... surprise surprise. For cars with gearboxes this definitely includes the revs as moving along the ideal driving force hyperbola certainly makes you faster. The driver does not have the time to check the exact number of the revs and remember when to shift. Instead it has been common practice to use different coloured LEDs, starting to light up in a fixed order when coming close to the shifting RPM. Ideally this would also take into account the current gear, the driver's reaction time and the shift time needed. Nothing new so far, F1 does it exactly like that and most FS cars also implement it that way.
The question remains how to actually display the revs:

  • Where to put the LEDs? Dash board (like RedBullRacing: F1 2013 Australia GP - Qualifying - Red Bull Racing - Sebastian Vettel) or steering wheel (like Lotus F1: F1 2013 Australia GP - Kimi Raikkonen Last lap Onboard - HD) or even in the helmet (Schumacher used a system like this for some time at Ferrari: Helmet-mounted shift LEDs)? There is no right or wrong answer in my opinion. Some argue that the LEDs are not easily visible/understandable when tilted/mounted on the steering wheel. But you do usually not accelerate/shift while having high steering wheel angles and in most installations the driver would hide the LEDs on the dashboard with his/her hands anyway. Additionally you need another ECU including connectors, housing etc., if you already have one in the steering wheel and want to put the LEDs on the dashboard. My personal opinion therefore would be to put them on the steering wheel. It saves the most valuable resources of a team: time and money and it delivers no performance advantage, if done in another way (mounted on the dashboard).
  • How to control the LEDs? All current F1 teams seem to light the LEDs up from left to right, while some FS teams let the LEDs light up from both ends such that the lights "join" in the middle. I think this is something which is highly depending on personal flavour. You might just want to think, whether it could confuse your driver if the LEDs light up against his/hers mother tongue's reading direction. For most of us it might seem normal to read from left to right, but Arabic for example is written and read from right to left. Therefore it might be more natural for a team from the middle east to have the LEDs light up from right to left.
The driver should also be aware, if fluids/parts of the car are too hot or too cold, of if there is a low or high pressure situation, so that he can act accordingly. Again, the driver does not need to know the exact number. It is sufficient, to be made aware of a problem to act accordingly. It also depends on how sophisticated your onboard systems are and whether you want them to "think" for the driver or not. For example the rev limiter could automatically reduce the limit in case of oil or water overtemperature situations, but you still want the driver to notice that. A pragmatic and easy approach is to just have single LEDs lighting up, if there is a problem. For example 3 LEDs, (water temperature, oil temperature and pressure) would do the job perfectly well. They could even be multi-colour (green, yellow and red). Positioning of these LEDs is not important as long as the driver is able to check them on the straights. (Low oil pressure indication should latch anyway otherwise the driver has to watch the LED in high G corners...you certainly do not want that).

An extra note about positioning LEDs: If you are planning to use high-power brightness LEDs, keep the following two things in mind:

  • Make sure that you have measures to reduce the brightness of the LEDs, if necessary, for example by connecting them to a PWM capable output/LED driver. Otherwise when running at dusk or in rainy conditions, the driver could be blinded by the LEDs.
  • The high-brightness comes usually at the cost of high directional sensitivity. So make sure that the central axes of the LEDs points towards the driver's eyes and not to his/her chest or even deeper. Otherwise it could happen that the driver is still not able to see the LEDs despite them being very bright.
The last information that could be beneficial is knowing the current gear. This is not necessary as the driver only needs to know whether to shift up or down. He should better be able to figure that one out on his/her own, but he wants to know at least whether the car is in neutral before letting go of the clutch. Therefore a single LED, indicating neutral, would do the job. For driver training a display showing the current gear can be helpful. F1 usually uses a single 7-segment-LED-display for this purpose as can be seen in the videos above. In my opinion this is by far the best solution: very good contrast, cheap and easy to control with a µC. For positioning please refer to the points given for the rpm LEDs, they apply to the gear display as well.

That is already it. The driver(!) only needs this information. All other information is only ballast for him/her and distracts from driving. But there are still the engineers.... as said before in a perfect world they have a downlink from the car and just watch all parameters on their laptops, but this is rarely the case for FS cars. Let's face it: even if such a system is in development or planned for the car, it does usually not work during testing or is unstable at times. Therefore there should be more information available for the people operating the car at the test track. I like how F1 does that, although they do not rely on it: They have two 4-digit 16-segment-displays and can change the shown values with switches on the steering wheel. These are again either mounted on the dashboard as seen here in Vettel's 2011 cockpit or on the steering wheel, for example Ferrari's 2013 steering wheel.

The pros of the 16-segment displays are, as mentioned before:
  • very good contrast, even in bright sunlight
  • low price
  • easy to control
  • you can also display text and let it scroll, if necessary
Some teams in FS use LC displays instead of 7-segment displays. This is also a good option, provided that it is a character display and provides sufficient contrast.

When it comes to displays being used, please avoid these common mistakes:
  • Do not use smartphones. Do not even think about it. It will look like a great idea at first, until you realize that they have to be charged/powered, usually do not provide good contrast, are pricey, need extra water-protection (unless you use a Defy) and consume lots of programming resources. Additionally they are hard to use with gloves. I am quite sure that from time to time a software engineering or design student will try to convince a team to use a smartphone in the steering wheel...
  • Do not use graphic displays unless you do not have anything else left to improve. I have seen it work and also done it myself for our '06 and '07 car, but remember: You have to control every single pixel. Even if you think that using nice symbols etc. on it would be a good idea: The driver has no time to enjoy them. If he has the time to look at them, he is not driving fast enough. Some design judges, especially those who have never driven an FS car, may like the bling-bling-factor of it, but that is it. So going down that road is a bad decision project management-wise in my opinion.
  • Don't design the display with detailed data with the driver as user in mind. The vibration and his/her concentration will usually not allow to take a look at it while driving. The display must only be readable, when the car is standing still for diagnosis purposes. These are two examples for a bad display design(the rpm LEDs seem to be ok though) in my opinion: example 1 and example 2. The driver will never be able to read the information provided while driving. (Also note that these are graphic displays, see my comment above).
  • Don't display any kind of boot or splash screens. I made this error once and showed our team logo on the display when the steering wheel ECU was reset/powered on. All drivers thought that the steering wheel had to boot and thus waited for the splash screen to disappear until they would try to start the car which was totally unnecessary.
  • Don't use any kind of operating system / steering wheel ECU that requires a boot time longer than 100ms. As said before, normal people are used to wait for a system to boot, so will the driver. It will just lose you time in Endurance, if your engine stalls and the restart causes a reset of the steering wheel ECU for example. The driver will wait for the steering wheel to boot up and every second boot time will cost you points. I am telling this, because when judging the electronics award in 2009 for example, there was a team with a Linux driven steering wheel. It provided a lot of nice functions and was even talking to the driver. However, it needed several seconds to boot and each time you switched it on, you had to wait a nagging amount of time prior to starting the car. It is just unnecessary in my opinion to use a full-grown OS for this task. BTW: With a steering wheel that needs to boot, this would not have been possible: Lewis Hamilton Steering Wheel Change Pitstop
A last word about graphic displays, before all teams using them go nuts: Yes, they look shiny. Yes, they can be designed to work and even act as a big gear display for example, but as said before: If you are able to invest the time of your sparkies elsewhere, then do it and skip the graphic/coloured displays and the price disadvantage. Performance and point gains lay elsewhere.

So to put it in a nutshell: I think F1 does a good job in visualising important data for the driver and is worth copying.

The second part of this topic, changing mechatronic parameters of the car, will be covered in my next post.

Please note:
If you think that any content in this post is worth discussing, please do so in the FSAE.com forum and not in the comment section of this blog, to make sure that everyone can benefit from the contents of the discussion, even if he/she is not reading this blog.

5 comments:

  1. Place your order within few clicks and get your dream car from company. The only online car selling company which is providing low cost cars and fast shipping options. reserve your car today before it is sold , the online stock is updated daily and thousands of cars are sold online.used Cars for sale

    ReplyDelete
  2. Vauxhall car dealers in Stafford have been massively damaged and there DVLA are quite a few used cars in Stafford that contain also been removed and that are awaiting the crush by the DVLA.

    ReplyDelete
  3. Superbly written article, if only all bloggers offered the same content as you, the internet would be a far better place..
    Japan auction cars

    ReplyDelete