WaterlooSPEED.com Progress Report

A New Year brings an update on the waterlooSPEED initiative!

National Instruments put together a video from NIWeek 2007 documenting the demo vehicle:

http://www.ni.com/compactrio/

Click on the “Videos: CompactRIO in Action” link. 3rd one down.

All of the project videos have now been released on YouTube. Check ‘em out:

http://youtube.com/profile?user=waterlooSPEED

The demo vehicle has been neatly stored for the cold Canadian winter ahead. Look for some major developments to be released in the spring.

Our lab space at the University of Waterloo has been vacated and we now seek garage space. Many thanks goes to Dr. John McPhee of Systems Design Engineering and Dean of Engineering, Adel Sedra for their support on this ambitious project and start-up.

Check back soon for future updates!

I wanted to make mention of a National Instruments conference that I recently attended in Munich, Germany called VIP Days (Virtual Instruments in Practise).

waterlooSPEED had a booth in academic section among some of the finest engineering schools in Germany. With posters outlining the project specifics and promotional videos being shown throughout the 2 day conference, it attracted much attention. This show has many more participants from the automotive sector than NIWeek in Austin, Texas. It helped that the prototype car is of German decent as well…

There were about 40 in attendance at the hour long technical session that I hosted even though I was presenting in English (the majority of the conference was in German). I exchanged contact information with some of the industry’s major players and made sure that project sponsors including WatCAR, Drivven and VW Waterloo received many plugs.

Thanks for all of your continued support! Please feel free to contact with any questions or comments.

During NIWeek 2007, I was interviewed by the VI Road Show team. They are a great bunch of guys that are passionate about demos powered by LabVIEW. Thanks for the feature!

http://www.viroadshow.com/2007/08/university-of-waterloo.html

We’ve been extremely busy since returning from NIWeek.

Due to overwhelming demand, I’d like to take a moment to post a link to our Introduction Presentation:

https://secure.niweek.com/2007/scheduler/catalog/catalog.jsp

Tech Session: TS1660

You will have to register on NI.com in order to download.

This presenation contains a good overview of the demo vehicle we built. It won’t satisfy the techie lust of all those engineers out there though…

Enjoy

*** Edit ***

It appears as though NI’s link to the presentation is broken. Hence we have hosted the presentation file on our server:

http://www.waterloospeed.com/NIweek07PPT.pdf

After reviewing the blog entries, I realise we made little to no mention of how the demo car’s engine runs on our control system…

The car starts, idles and revs very similarly to when it is being by the Motronic ECU. I drive the car with the waterlooSPEED controller. As with any project of this magnitude, there is still some tuning that needs to be done. I’m the sort of guy that is never quite satisfied with my projects. There is always room for a little bit more…

The biggest challenge to date has been dealing with transient response of all sorts. I don’t want to get into the real nitty gritty details but when attempting to crank an engine over, the initial solenoid actuation of the starter motor draws a huge amount of current (voltage drop). We struggles with power to the modules, controller, sensors and such. Acquiring a clear crank signal for syncronization during the transition from stalled crank to cranking speed is also tricky. Once the engine was idling (steady state), dipping into the throttle creates another complex control state.

With transients out of the way, the speed density formulation works quite well for calculating fuel mass. We need to again take a look at calibration data.

The big “show and tell” that we have been longing for is just around the corner (ie. next week).

We drove the Jetta to a special shipping agent yesterday in Alymer. They specialize in transportation from Toronto to Houstin so Austin is isn’t far out of their way. Photos of the vehicle loading should be up soon.

Unfortunately, waterlooSPEED will not be explicitly featured in the keynote due to a number of scheduling and ownership issues. NI has promised the final video compiled by Michael Litt showcasing the construction of the Jetta demo from beginning to end will be shown. This viewing will occur on Thursday morning during the Academic portion of the keynote.

It is my hope that I am able to network at NIWeek and find some potential customers to put NI and Drivven data acquisition and control systems together for. It wil also be great to get the exposure that the University of Waterloo deserves on the international level. The Systems Design curriculum that we are a product of is an excellent fit to the “Graphical Systems Desgin” theme of NIWeek 2007.

Chris did an excellent job designing and constructing a very robust method for us to “three way connect” the ECU and our cRIO based controller. It has worked flawlessly up until now…

A position sensor for the accelerator pedal is reporting a signal out of the expected range to the factory Motronic ECU. Accordingly, the ECU commands “limp home” mode where throttle is limited to 10%.

After many late nights in the blistering Canadian summer heat, the wire bundle has been identified. A temporary fix allows the signal to remain in the safe operating range until post NIWeek where the wiring will be properly replaced. With 121 pins on this engine computer, sorting and identifing wiring can be a real challenge.

Whenever I have had a spare hour or so, I have been taking the time to create a more concrete labelling system for the waterlooSPEED demo. Redundent wires have been removed the “birds nest” has been sorted by twisting and secured wires together. The result allows for much fast troubleshooting and is more pleasing to the eye.

I appologize for not keeping up to speed with blogging but every spare moment I have had since convocation has been spent on sorting technical details to get our demo vehicle ready for NIWeek 2007.

Rishi and Devin have been kind enough to take breif breaks from their worldwide travel to stop by the laboratory and give me a hand. Rishi’s networking skills have come in handy and he has successfully made the wireless communication with out PDA active. The prototype vehicle’s pneumatic suspension can be actuated from very far away (802.11 Wireless specification). Not only can the suspension be actuated but air pressure in each airbag is reported as well.

The PDA’s front panel is seperated into three tabs. One is controls connection, the second air suspension and the third is throttle body control. In other words, from 30 meters away the user is able to “rev” the motor. We realise tehis really doesn’t hav much use unless you are James Bond but it puts smiles on engineering faces… In other words its cool.

waterlooSPEED has been featured in the University of Waterloo’s Daily Bulletin.

The article talks of the NIWeek feature coming up next week. Expect more media outlets to come…

http://www.bulletin.uwaterloo.ca/2007/jul/31tu.html

It’s been a while since the last update, but there has been a lot of progress in that time.

In it’s current state, the controller:

• reads all sensor inputs (IAT, ECT, MAP, RPM, spark, fuel injection, O2, throttle position, pedal position, cam and crank signals). All sensors have been properly calibrated and verified against VAGCOM measurements
• controls the throttle based on pedal, idle speed, and engine status
• uses speed density calculations to compute fuel injection timings
• determines spark timings based on our reverse engineering spark advance and dwell time maps
• syncs with engine position for precise timing events

We have successfully tested the controller in a number of situations:

• raised and lowered the pneumatic suspension of each tire
• accurately controlled throttle body position based on pedal position both in the lab and on the vehicle
• pulled a fuel injector and monitored fuel injection while cranking
• pulled a spark plug and monitored sparking while cranking
• verified fuel and spark timing signals (position and dwell) against stock ECU signals at idle and while revving the engine

With performance verified by the above tests, we were ready to try cranking while completely controlling all engine operation. While going through our pre-cranking checklist, we discovered that we were not receiving throttle position information. After extensive testing, it appears that the Drivven throttle body module is no longer functioning properly. At present time, it is unclear how or when the module was damaged, but we are verifying all of our wiring to ensure this does not happen again. We are looking into acquiring another TB module so that development can continue, and we can begin to tune our performance.