It will be a very long time before there is demand for autonomous cars in markets like Nigeria (or in the nearer term, Turkey, Brazil, India and China) simply because cost of driver labour is so low and utilisation of transport is so high.
Most vehicles are already effectively informally time-shared along extended family/community/business lines so they are very highly utilised unlike a a suburban American family car.
And drivers are very low cost – I can pay a guy to drive me thousands of kilometres across Himachal Pradesh and Ladakh using his vehicle and fuel for a few weeks for less than what a San Francisco native pays to fill his car with premium unleaded.
In these markets the cost of the driver is vastly less than the 30% of total vehicle operating costs that is motivating Uber and Waymo.
I’m driving myself through northern Italy this week and I would suggest most autonomous systems would do a better job of not damaging my rental car in narrow cobblestone streets than I would.
Most of the road infrastructure in Europe, North America, Japan, Singapore, Hong Kong and the Commonwealth is now accurately mapped and only a very small proportion of it is changed from year to year, always by authorities which would be able to update the cloud autonomy databases. If autonomy works, these are the markets to focus on dominating – Nigeria will be OK ;-)
There will definitely be cultural barriers in many of these markets where car ownership and driver skill are significant components in the expression of male identity. If I were an autonomy strategist that is the single most difficult barrier to adoption that I’d want to invest heavily in.
How do I persuade an American SUV or pickup owner that paying more for a version with autonomy isn’t going to emasculate him? How do I ensure a German, Italian, Spanish, English or Australian owner of a performance car that it’s OK to pay more for the version that includes autonomous mode?
Look at the global audience for the TV series Top Gear! The world has a lot of car nuts (I should know, I’m one!) and converting them to autonomy is a far greater challenge than developing the AI and ML.