Facebook link Linkedin link Soundcloud link RSS feed

Back to the ’90s?

25 March, 2026 in Cars, Computers, Games

My passion for human-in-the-loop vehicle simulation was ignited in the late ’80s / early ’90s by two incredible Atari coin-op games: Hard Drivin’ and Race Drivin’.

For the first time, it was no more a simple “steer left / steer right” matter: Atari introduced a full vehicle simulation, where the challenge was to reproduce real driving manoeuvres to keep the car on track—whether in stunts or at high speed.

I spent countless hours playing those games. And they also inspired me to start a personal project in the early ’90s on the mighty Commodore Amiga, codenamed Impossible Drivin’ (yes… you can probably guess the reference)!

It was my first attempt at simulating a real vehicle in real time. It never became a full product and was never officially released. It featured a simple open-wheel car driving on a basic polygonal track, with just one goal: stay in control at high speed.

Despite its simplicity, Impossible Drivin’ laid the foundations for the following AVC simulation library and the Virtual Grand Prix series, implementing my early vehicle simulation algorithms. Even if, by modern standards, those algorithms were very limited, they were anyway able to capture the essence of fast driving. The model simulated the different behaviour of front and rear tyres, when pushed toward the limits of their friction circles depending on user input. The result was surprisingly fun: the car could easily understeer or oversteer, and the player had to react with the correct manoeuvres to control the vehicle and stay on track.

Even with such simple algorithms, it felt “realistic enough” to be genuinely fun to drive. In fact, my friends and I spent hours playing that early demo, even though it was nothing more than just a minimal sandbox experience.

A few weeks ago, while experimenting with the latest version of the Godot engine, I had a thought: what if I disabled the built-in physics and reimplemented those old Impossible Drivin’ algorithms?

Godot already provides so many features out of the box that recreating the project would be straightforward: just rewrite those few lines of C code in GDScript for the car physics, and let the engine handle everything else (3D, input, etc.).

Well… here it is: the IDFG Project — Impossible Drivin’ for Godot.

And yes, the goal was fully achieved.

The car is extremely fun to drive. I would even say it feels more enjoyable than many modern, full-featured simulations.

Driving in IDFG brings back those ’90s vibes—when simulators were realistic enough to reward good driving, but not so extreme that they became frustrating or intimidating.

Instead of tweaking endless parameters, you just jump in and enjoy throwing the car into corners.

So now the question is:

what’s next?

Should this remain a simple demo, or is it worth turning into a full game with those retro vibes?


The 86 Tribute

17 May, 2024 in Cars, Computers, Games

The “86 Tribute” series for the VGP3 Engine is my heartfelt tribute to an iconic sports car renowned for its exhilarating driving experience; with its lightweight design, rear-wheel drive, low center of gravity, and perfectly balanced chassis, it has captured the hearts of driving enthusiasts around the world.

As a token of appreciation for all the joy and excitement this little japanese sports car has brought into my life, I’ve created this mod to share the experience with fellow enthusiasts.

It is my humble tribute to one of the most fun cars I have ever owned. If you want to have fun driving a “virtual” version of the GT86, you can download (for free) a special version of the VGP3 Engine, containing that series, from here:

https://www.virtualgp.info/index.php?page=download


Tyre-road heat transfer

3 November, 2023 in Cars, Computers, Games, Science

Tyre-road heat exchange

I have been studying vehicle dynamics for almost thirty years, since I set out to create my first driving simulator, Impossible Drivin’, for the Amiga platform. Over the years, I have primarily studied the complexity of the interaction between rubber and asphalt, which is responsible for 90% of the quality of the simulation, and I have developed several algorithms to effectively implement the thermomechanics of the polymer.

In this scientific paper, I publish one of the formulas I developed (and effectively used in my AVC simulation library) for the dynamic calculation of heat transfer between rubber and asphalt, a formula that replaces the simple constant Hc normally used in the literature:

https://www.mdpi.com/2076-3417/13/21/11996


GT86 Fun

5 May, 2017 in Cars

GT86 car

I like my GT86.

While the GT86 isn’t the most powerful car in her class, I like how the engine begs to be revved, and how she offers “just enough” performance for my needs and driving style. The lightweight body and well balanced rear-wheel-drive layout make her a very good handling sports car; yet, she’s not something “extreme”, the GT86 can easily be used as an everyday car!

All in all, I’m very happy with her!


To pull or not to pull

8 January, 2015 in Cars

In the last months many words have been spent about the bad handling of Ferrari’s F1 car (the F14T project). Debate revolved mainly about the unusual pull-rod front suspension scheme.

Now, there’s a lot of speculation about the new Ferrari F1 project, about this topic: a pull-rod scheme will be used again for the 2015 car?

In my opinion, the problem of the F14T has never been the push or pull-rod scheme. The problem is with the configuration of the double-wishbone.

For aerodynamic reasons the attachment points of Ferrari front wishbones are very high; this solution gives an highly “sloping” configuration of the suspension arms, a thing that is immediately noticeable from a front picture of the car. Such a solution makes the front tyre, when the suspension moves, travel not only in the vertical direction, but also laterally; in other words, whenever there is a pitch movement of the car nose, the wheel track changes.

Of course, too much lateral movements of front tyres increase tyre wear, because additional lateral slip is added; so, this suspension may work only if combined with very stiff setups. And a stiff front setup means, usually, understeer.

So, the real problem about the new Ferrari F1 project is not whether it will sport again a pull-rod scheme or not; the real problem is that, unless the front suspension is heavily modified (attachment points etc.), the car will be again heavily understeering; and given the fact that both Ferrari drivers, Kimi Raikkonen and Sebastian Vettel, don’t like at all an understeering car, that will be really a problem!

It seems, from rumors, that James Allison, the new Ferrari technical director, is making changes to the original project (as designed by Tombazis), especially in the front part of the car; let’s hope that such changes will positively affect the new Ferrari F1 car handling!