Testing, testing

Ironically, when I found myself at the wheel of the same kind of prototypes that had so interested me as a youngster, it wasn’t the realisation of a lifelong dream achieved by dogged determination and a refusal to take no for an answer, but simple, dumb luck. Which is a bit anticlimactic, I know. In my head, development engineers had to be a special breed and I never even considered myself to be cut from that kind of cloth. I had always assumed that, like airline pilots, you would have to complete thousands of hours of training, racing cars on tracks all around the world or be born with some supernatural driving talent to be part of that special club.

Not so. As I found out in my first week as a validation and integration graduate, they actually trust run-of-the-mill engineers to get on with the job. No need to crash land your spacecraft into planet Earth as a baby or be bitten by some sort of radioactive spider – you needn’t be a superhero at all. Actually, you get some level of training as part of the role, but this will also very much depend on what sort of cars and testing you do.

There are different levels of development cars. For example, I used to drive mostly pre-series vehicles, which were usually more mature and better put together than early test mules and, for the most part, closely resembled final production designs under all that disguise.

Unlike mules or prototype cars, which were built in smaller workshops by the early development team, these were assembled with the current production cars on the line and also used as a way to trial out manufacturing processes before launch. With each build phase, validation engineers like me got a chance to assess different versions of the software and hardware. Early test cars usually ran on very basic calibrations and they could feel clunky – most of the bells and whistles would still be missing at that stage – but they matured with every iteration to get closer and closer to the final product that our customers would one day drive on the road.

For this reason, testing ought to be as relevant to the actual final intended use of the car as possible. There’s a common misconception that all testing happens in exotic locations or by driving fast around a track. And while some of my colleagues in the vehicle dynamics department did often spend time at the Nürburgring, skidding on icy lakes in Sweden or dune-busting in the Middle East, in reality the majority of testing is somewhat more mundane.

In the powertrain world, for example, engines can be individually run for thousands of miles in a test cell within very accurately controlled environments and with very little human interaction. So, when you get it into a car, it’s about proving the integration with the chassis, transmission and everything else around it. A lot of the standard test drive checks – usually done on the road or dedicated test facilities like Millbrook Proving Ground in Bedfordshire – include fairly common sense tick-boxing exercises, like checking top speed, testing the different start-stop triggers, or ensuring both engine and gearbox behave as they should under all types of load, roads and gradients. However, test engineers also have to get creative.

As David Twohig very wittily observed in his article Used and Abused, cars are designed with ‘reasonably foreseeable’ customer use in mind. This means test engineers will try to come up with the most random scenarios or sequences of events that a customer could get himself into (intentionally or not) and pray that our development colleagues had foreseen precisely that.

I’m not going to lie; the first time I tried to shift an automatic gearbox from drive to reverse at speed I gritted my teeth waiting for an expensive metal-on-metal symphony of destruction as the car tried to go backwards while still going forwards – but nothing of the sort happened. The car just continued about its business, unflustered.

I had to summon similar courage when I tried to disembark from another automatic car while it was still moving. I’d set it into drive and waited for it to start creeping forward before undoing my seatbelt and opening the door. But once more it just stopped dead in its tracks before I could even dangle a foot out the door.

It’s not always that dramatic either: sometimes it’s just about channelling your inner toddler and pressing a dozen buttons and waiting to see if the dash lights up with all kinds of warning lights. Either way, please don’t try any of this at home – just be assured that whatever you can think of will already have been anticipated at some point and made foolproof by a brave team of testing engineers. At least I hope so.