FEATURE // Are autonomous cars ready for Africa?
As autonomous vehicles reach real-world roads, African cities face the challenge of navigating complex, unpredictable environments
If you happen to have found yourself on a street in downtown San Francisco any time in the last two years, you will know that self-driving cars are no longer an experimental technology, and certainly not the transportation of the future. As we speak, fleets of self-driving Waymos are embedding themselves into the infrastructure of today in cities around the world. But as these autonomous vehicles move from controlled pilots to commercial services in Tokyo, Madrid, Abu Dhabi, and Shanghai, some may wonder if these models will work in Lagos, Johannesburg, Cape Town, or Nairobi.
African cities have different road systems, transport economics, informal mobility patterns, and infrastructure constraints. In South Africa, for example, an enormous minibus-taxi system already provides flexible, demand-responsive transport for many commuters. Still, African-founded companies are finding a place in the global autonomous-vehicle (AV) ecosystem. Moove, the transportation startup founded in Lagos in 2020, began by financing vehicles for drivers who struggled to access conventional credit. That business has grown into a global mobility platform operating approximately 42 000 vehicles across 29 cities in 13 countries. Moove has since moved into autonomous-mobility infrastructure, partnering with Waymo to manage robotaxi fleet operations and investing in the systems needed to maintain, charge, and deploy driverless vehicles at scale. In August 2026, it raised USD250 million at a USD2.1 billion valuation to accelerate that strategy. The company’s trajectory shows how an African-founded business can enter the global AV supply chain. And yet, Moove's exit from Nigeria in October 2026 underscores how challenging the economics of mobility can be in individual African markets. So, what will it take to make AVs work in African cities? We look to industry experts for some clues.
Autonomy is here, but not all cities are the same
The International Energy Agency says Level 4 electric driverless taxis are already operating commercially in more than 20 cities worldwide, while fully autonomous Level 5 vehicles remain a more distant prospect. In 2025, the global robotaxi fleet more than doubled to around 8000 vehicles, but commercial deployment remains concentrated in China and the United States, with early services emerging in the Gulf while Africa remains largely outside the activity map.
A 2025 World Economic Forum report paints a slower picture of the AV rollout, which points to technological, regulatory, and economic complexities that continue to hold back large-scale deployment. At city level, operators must map and adapt their software to each city’s unique driving environment.
We spoke to Professor Marianne Vanderschuren, a transport planning and engineering professor at the University of Cape Town, about what makes an AV suitable for an African city, rather than importing technology developed for a US city. Prof Vanderschuren recalls first encountering automated vehicles in 1996. She says it has taken until now for the technology to enter the market partly because the technology has improved, and because international legal and liability issues have been addressed. In theory, AV can function anywhere; in practice, African environments are more complex and less structured, with road conditions and unpredictable behaviour from other road users presenting barriers to deployment.
Volvo South Africa notes that today’s AV technologies are not designed for African road environments out of the box. They rely heavily on structured conditions, including visible lane markings, predictable traffic rules, reliable positioning, and high-definition 3D maps. When autonomous-driving technology is introduced into African cities, it may encounter conditions it was not designed for. In informal neighbourhoods, roads can be narrow and unmarked, street layouts can change, and traffic often reflects local transport practices rather than formal road planning. A 2026 commentary in Nature Cities, drawing on examples from Kampala and Nairobi, argues that preparing for autonomous mobility will require more than importing the technology: communities and local transport operators should help map streets, improve signage, and shape solutions around how people move.
A path for the AVs
Vusi Machie, Head of Product, Planning & Pricing at Volvo South Africa tells us that this does not mean that AVs cannot work in Africa, but software trained on clean Western suburbs cannot simply be imported and expected to handle local roads safely. A realistic deployment strategy would start in controlled, predictable environments, such as mining operations, ports, industrial corridors, or dedicated bus lines where routes, maintenance, and traffic patterns can be tightly managed before moving into busy city centres.
Machie suggests treating autonomy as shared public transport rather than selling individual driverless cars. This would mean:
- Feeder shuttles over private cars: Prioritise fixed-route, high-capacity shuttles integrated into existing public transport, including matatus, BRT feeders, and informal bus routes.
- Smart roadside infrastructure: Where visual cues are unreliable, connected roadside sensors (V2X) could help vehicles locate themselves rather than relying entirely on onboard cameras.
- On-vehicle edge computing: Vehicles would need perception software that can operate offline or with weak connectivity, handling local surprises without stable cloud access.
- Targeted corridors: Rather than mapping an entire city, municipalities could start with high-density freight or passenger routes and scale outward.
Meanwhile, the Transport Safety Laboratory at South Africa’s Council for Scientific and Industrial Research (CSIR) is building an evidence base around the human, infrastructure, and vehicle factors that autonomous systems would need to navigate in South Africa, using real-world road data alongside simulation. A March 2026 post from the CSIR, entitled "Scientists on the scene as road deaths climb", strengthens the safety argument for AVs, highlighting work on advanced driver assistance systems, automatic braking, and vehicle technologies tested against South Africa’s road conditions and driver behaviour.
The challenge of integrating AVs into African cities
An AV entering an African city would have to navigate a transport system already shaped by private cars, minibus taxis, buses, trains, walking, and cycling. In Cape Town, the City’s 2024/25 estimates show that private vehicles account for 55.2% of commuting trips, while minibus taxis account for 22.5% – around 945 000 trips. Any move towards autonomous mobility would need to consider how new services could complement existing transport, rather than simply replace it.
Prof Vanderschuren argues that South Africa presents a particularly challenging environment for AVs because driver behaviour is more complex and does not follow a normal distribution. Minibus taxis, pedestrians, and other road users may behave in ways that automated systems trained in more predictable environments struggle to anticipate. In other African cities, large numbers of motorised two-wheelers add another layer of complexity. These challenges are not necessarily insurmountable, she says, but they require research and development tailored to local conditions. She is not currently aware of anyone undertaking this work at scale in South Africa or across Africa.
Given the continent’s road-safety challenges, Prof Vanderschuren argues that safety should come first, with early trials potentially taking place on dedicated infrastructure or in controlled environments. But autonomous mobility is not limited to replacing conventional public transport with driverless buses or robotaxis. Asked how automation might affect transport-sector livelihoods, she emphasises the need to account for the many people whose incomes depend on existing transport services.
For now, she sees first- and last-mile services as a more practical starting point. These are short journeys that connect people from their homes to a bus or train station or take them from a station to their destination. Small autonomous shuttles or on-demand services could potentially fill gaps in existing public transport networks, although their suitability would depend on local safety, costs, and demand.
Who benefits from robotaxis?
AVs could improve road capacity by allowing connected vehicles to travel in closely spaced “road trains”, moving more traffic through the same space while reducing crashes. One of Prof Vanderschuren’s studies from the late 1990s estimated that vehicle throughput could increase by around 30% without changing road geometry. More recent research suggests the effect depends on how many AVs are on the road and how they interact with human-driven traffic.
Prof Vanderschuren expects wealthier neighbourhoods to be served first, raising an equity concern, but argues that robotaxis would not fundamentally change where people can travel as private cars, robotaxis, and public transport would continue to serve the same city. Over time, larger robotaxi fleets could reduce waiting times for public transport users, but their impact on congestion and employment would still need to be considered, particularly if robotaxis are operated as private car-sized vehicles.
A 2025 review of 608 studies found that smart mobility research in Africa has focussed heavily on technological solutions, with less attention to users, city-wide systems, and the socioeconomic effects of new mobility. PwC’s South African Infrastructure Outlook similarly argues for transport systems built around people rather than cars, including public transport, active mobility, and demand-responsive services.
Scaling autonomy
Prof Vanderschuren emphasises that technology is only part of the challenge in scaling AVs. Onboard sensors can handle much of a vehicle’s immediate surroundings without a network connection, but reliable connectivity is still needed for communicating with passengers and navigating routes using GPS. Charging presents a separate challenge, particularly for electric fleets, which require adequate charging infrastructure. Her department is also working with South Africa’s Presidential Climate Committee and the Western Cape government on how to roll out new energy vehicles in South Africa and the wider African context.
Machie points out that South Africa’s biggest challenge is building the ecosystem needed to operate AVs safely and at scale. This includes closing policy gaps, improving road and mapping infrastructure, and developing the local skills needed to maintain autonomous fleets.
South Africa’s National Land Transport Strategic Framework recognises AV technology as an emerging technology that needs to be incorporated into the country’s integrated transport system. It also notes that, at the time of publication of the report, South Africa had no transport legislation specifically regulating the deployment of new technologies such as autonomous and electric vehicles.
For African cities more broadly, the lesson is that autonomous mobility will need to be designed around the realities of the cities it is meant to serve, and innovators like Moove are showing us what that might look like. By building the fleet operations, depots, and charging infrastructure that AVs will need, Moove offers one glimpse of how the transition from shared mobility to driverless transport could unfold. Whether that model can be adapted to African cities remains to be seen.
by Denise Mhlanga
