Tablet-sized controller to manage autonomous motoring

A name like 'zFAS' for a sophisticated new control system that will guide cars around the world's streets and byways seems rather humble.

Audi, the company backing the development of the new system, could have opted for 'Colossus', or 'Skynet' instead.

The new system, which is scheduled to enter production within two years, is described by the prestige manufacturer as delivering "high-end computing power on a surface the size of a tablet PC. Joining Audi in the development of the new device is known expert in its field, Delphi – the team-leading project partner – ably assisted by TTTech, Mobileye and nVidia.

According to Audi, in a press release issued last week, zFAS (it's a German acronym apparently meaning 'central driver assistance controller') is unlike other systems under development by rival firms, because it is fully integrated "in a central domain architecture", rather than relying on "spatially separated controllers".

Despite being based around that concept of the "central domain architecture", it's also modular, making it "future-proof", Audi claims. That means it's flexible and scalable – so it can be retrofitted with new software and hardware to deal with increasingly complex situations in the future. The groundwork for the system is there, in other words. It's like upgrading RAM, CPU and OS on your motherboard and hard drive.

Audi has done the maths and claims that the zFAS has computing power equivalent to a car's entire electronics system – which includes all the gizmos to operate your multi-zone climate control, Bluetooth audio streaming, keyless entry, engine management, electronic dampers, stability control, airbag deployment, collision avoidance systems... everything.

But zFAS is ultimately just the brains. Companies such as Bosch, Continental, Valeo and Delphi will also be supporting the effort with the car's nervous system: sensors and actuators for braking, steering and drivetrain operating parameters.

Further into the future, Audi plans for its piloted cars to learn and adapt from past scenarios. Data from zFAS will be uploaded by mobile phone network to a 'cloud' repository where it can be analysed and, if necessary incorporated in new protocols for dealing with specific situations. Those protocols will then be downloaded into the cars on the fly.