In Summary
  • It is therefore becoming more and more necessary to insure one’s car, but also in the interest of national economics, it is becoming more and more necessary not to crash one’s car.
  • One way of ensuring you don’t crash is by building advanced cars, which is what makes them horribly expensive to fix when they do crash.

  • Another way is by driving better, taking more consideration on the road, and with a reward scheme in favour of the latter approach, this may just come to happen.

Part 1: Big Brother is Watching

It is easy to define good driving: fuss-free, impact-free, smooth, courteous and law-abiding. However, evidential material in support of or contrary to the fact is hard to come by. As we saw last week, it’s fairly easy for respondents to lie through their teeth. To add to this, the human mind is cryptically self-indulgent so it may not be a reliable data source to ask respondents to diagnose themselves. They will lie, they just won’t know they are lying.

But is it really that hard to prove good driving or the lack thereof? Before we answer this, let us backtrack a little.

Social media engagements on this subject reveal a general consensus that insurance premiums should in fact be pegged on driving skill and record, but there is the widespread belief that “good driving” entails a track record devoid of citations and crashes. This is not comprehensive (pun intended) because it is possible to evade bookings and crashes but still drive like someone with a single-digit IQ. I have seen people get by on a combination of tolerance from fellow drivers and blind luck, and they will insist they are good drivers because they have had no legal entanglements with traffic police or physical entanglements with other cars, but their driving styles can best be summed up in the question: What the heck do you think you are doing?

Usually, it is only a matter of time before the universe rights this wrong and they end up in a massive shunt that decommissions them and saves the rest of humanity from one more bad driver. I know this sounds callous, but one can only push the limits of probability for so long before the dice comes up deuces and the game is over.

Insurance companies base their operations around the word “risk”, but after my recent perusal of some intellectual property by one enterprising company, it is easy to see that they have been slightly on the wrong. Most of the risk assessments are either based on assumptions that could be erroneous (I have always been a very good driver from an early age but the insurance companies were not going to buy that, if I made such a claim 15 years ago — alternatively, someone who has driven for 30 years is not necessarily a flawless helmsman) or on data that is inconclusive: if you have several traffic offences to your name or have sent a large number of motor vehicles out of circulation, definitely you are a high-risk client, which makes sense, but what about those inept drivers who are surviving by clinging to the hems of Lady Luck and their number is not up (yet)? These too are high-risk clientele, and these are the ones you need to watch out for, but you can’t just tell on the face of it.

Or can you?

Enter Autocorrect, and its potential to be shrouded in a cloud of controversy. The name may sound a bit unimaginative and plagiarised from a completely unrelated context, but if you think about it, it’s actually quite clever. Auto- for automobile and -correct for rectification. Car-related rehabilitation, and who better to wield the stick than the folks who got your back when it all goes down the latrine?

Autocorrect is a product based on the real-time analysis of your driving style, such that the better you drive, the lower your premiums, but try any track-esque manoeuvres in a housing estate and it will cost you when you next renew your cover. You are rewarded for driving like you were actually taught and penalised for putting yourself and others at risk.

This is how it works: a telemetric device is installed in your car. This device records all the instances of abnormal acceleration, harsh braking and aggressive cornering, or in other words, the subset of circumstances that typically precede an accident (and thus expensive claims). It is normal to have a certain number of such “incidents” (the data sheet called them incidents), but if your quota is higher than others then that means you are a risk. That means they can tell you are a bad driver even though you have spent your entire driving life dodging the consequences. The blurb alleges that the device records temperatures and charging as well, though for the sake of pertinence, we will gloss over that.

These incidents are logged over the duration of your cover and come crunch time, you will be presented with proof of your indiscretions and your invoices adjusted accordingly forthwith and or until you show improvement behind the wheel.

Now to settle a few questions I’m sure you already have. Heritage Insurance Company, the innovators behind Autocorrect, insist the installed device is not part of tracking equipment. It will record what the vehicle is doing and when it did it, but not where. Worth noting is that the device is voluntary.

They say if you accept this service and drive like you should, you could see — among other things not important to this conversation — a drop in your premiums to 15 per cent.

*Addendum: I will not delve into the exact workings of the system since that information can be found by a bit of searching online and there is a lot to cover against word count limits.

Part 2: Crunching the Numbers

If you have an aversion to any kind of math or scientific research, feel free to skip this section, though it is mercifully short. Anyway:

The telemetric device logs the driving style (acceleration, braking and cornering), the distance driven (driving farther statistically increases chances of a crash for obvious reasons) and the time of driving (night driving carries more risk, where “night” is defined as “the time between 10pm and 4am). Your driving score is based on a combination of these three parameters. However, in what looks like a glaring oversight, speed is not a factor.

Why not? Speeding is a major contributor to traffic incidents, mostly because most drivers lack the skills and characteristics necessary to handle a vehicle at high velocity. Many people crash specifically because they were going too fast for the circumstances and failed to stop or turn properly. I find this odd.

And now the numbers.

1. In a survey done with a sample size fairly evenly split between male and female (53 per cent vs. 47 per cent), it turns out women scored higher, with an average score of 85, while men trailed them at 83 per cent. Ha! So perhaps “lady-driven” actually is a valid selling point when disposing of a used car.

2. As far as brands went, Mercedes-Benz led with a score of 89, closely followed by Land Rover at 88. The (in) famous Subaru brand came in ahead of Honda (81 compared to Honda’s 88) but was narrowly pipped by Toyota, which stood at 82. Mazda led them both at 83. I try not to judge people based on the car they drive, but this may be a bit telling. However, even more telling is the technological advancement of the car in terms of driver assistance. You may think that driver assistance would raise the score, but I dare say they would reduce it since factors like adaptive cruise control, lane keep assist, automatic emergency braking and whatnot would cause harder acceleration, braking and yaw since they are mostly reactive and override the driver input where necessary. That is a whole other discussion.

3. People drove the farthest on Saturdays, and the least on Sundays. Oddly enough, while Sunday garnered the best score of any day of the week (drive like a Christian, James May said), Friday easily took the worst. Drink-driving perhaps? Or fatigue after a long week of work?

Part 3: Road Safety

The whole Autocorrect concept may tickle my fancy given my educational background, shamelessly biased towards mathematics and science, but what really got me interested was the implications it may have on my pet subject: road safety.

During my training in Baltimore last year, part of the discussions revolved around how to incentivise drivers to behave better on the road. One of the first ideas thrown down was a subsidisation system by the government, such that insurance premiums for good drivers be lowered or heavily discounted. Fast-forward one year later and this sentiment was echoed by NTSA Director-General Francis Meja at a road safety forum that he and I attended as guest speakers at Catholic University of Eastern Africa, an event organised by the HOG Foundation (look it up).

(Clarification: As I mentioned, these are IDEAS that are being bandied about, and the fact that some of them came from a high-ranking civil servant does not make them policy, unless otherwise ratified. Don’t run to your insurance company demanding discounts because “the NTSA said so”. The NTSA never said so.)

Well, if this idea is to come to fruition then something like Autocorrect would come in very handy, wouldn’t it? But, if you think that is the end of that, then you clearly underestimate the depth of my training by the WHO in the US.

Road safety is still a thorn in our flesh, as could be seen from the recent spate of crashes involving high-profile individuals. To combat this menace, a sane and sober approach is needed, not quite the prayer and knee-jerk bawling that usually follows these events (I’m sorry, Kenyans, but deriding the safety standards of a Nissan Navara is not going to undo the Thika Road tragedy). We need a calculated approach, in all meanings of the word. We need an agenda that is evidence-based and data-driven, because these are what will guide the thought leaders and policymakers on what direction to channel their energies.


Cars keep getting more and more complicated as they advance, to the point where an American survey established that the repair cost of a corner-damaged Kia K900 would come to about $34,000, which is crazier than what we saw last week. Even crazier is the fact that the Kia K900 retails for $50,000. The industry average? $8,000. Men lie, women lie, numbers don’t. It is therefore becoming more and more necessary to insure one’s car, but also in the interest of national economics, it is becoming more and more necessary not to crash one’s car. One way of ensuring you don’t crash is by building advanced cars, which is what makes them horribly expensive to fix when they do crash. Another way is by driving better, taking more consideration on the road, and with a reward scheme in favour of the latter approach, this may just come to happen. It will at least put a kibosh on the vicious cycle that is the overcharging overpaying dance between the insurance company, the garages contracted by these companies and the customer who shuttles between both.

Having car trouble? Email