Kenya is still a long way from being a leader in Africa’s electric vehicle (EV) adoption, but despite the slow pace, the use of artificial intelligence (AI) is accelerating as this transformation takes shape.
In this episode, we dive into the real-world applications of artificial intelligence across the EV value chain—how it’s being used to optimize fleet operations, improve battery performance, manage charging infrastructure, and create predictive maintenance systems.
we’re joined by Warren Ondanje, Managing Director of the Africa E-Mobility Alliance (AfEMA), to unpack how AI is shaping Kenya’s electric mobility landscape where he shares insights from his work building tech platforms that support EV deployment in Kenya, offering a glimpse into how local startups are innovating at the intersection of mobility and data science.
The conversation also explores Kenya’s broader transportation ecosystem, including the infrastructure gaps. We discuss how policy can catch up with innovation and what an AI-driven, low-carbon future actually looks like on African roads.
Whether you’re a tech enthusiast, climate advocate, policymaker, or entrepreneur, this episode offers a grounded, forward-looking view of how smart technology can make clean transport work for Africa.
Show Notes:
- W.O: Artificial Intelligence and Africa’s Electric Mobility: Insights, Investment & Innovation by Warren Ondaje
- Business Daily: Ensure sufficient power to sustain future AI, electric vehicles demand by George Wachira
- Energy for Growth Hub: Three Steps to Unlock the Potential of AI in Africa by Rose Mutiso
About Ondaje:
From a university dreamer straddling between engineering, radio production, and AIESEC events to a dynamo powering Africa’s e-mobility revolution, Warren Ondanje is all about responsible business and brand. Fast-forward, and he’s the Managing Director of the Africa E-Mobility Alliance, snagging a $150,000 ClimateWorks grant for EV battery circularity, turning the Africa E-Mobility Week into a 300+ delegate hotspot, and launching the continent’s biggest E-Mobility Data Portal—tracking 175 companies and $370 million in investments. Before AfEMA, Warren was at ARC Ride Kenya where he built a 20+ member crew in 2021, rolled out the initial pilot of 30 electric motorcycles, left a blueprint for rolling out another 300 electric motorcycles, and spearheaded duty reduction initiatives through local assembly to the tune of 30%—responsible businesses that lead to better customer offerings.
Transcript:
Njenga Hakeenah: Kenya has emerged as a leader in Africa’s EV adoption. Since AI is playing a crucial role in accelerating this transformation in other countries, how exactly is AI shaping the future of mobility in Kenya?
Countries like the US, China and Germany are among those that have incorporated AI into their electric vehicles with China’s.
Seven months ago, the Geely Ji Yue 01, the world’s first passenger car with level 4 autonomous driving, was launched. The autonomous driving levels are rated from zero where the car has to be manually operated by the driver to level five meaning that a car can drive completely by itself without any human interaction.
However, today we’re not discussing the Geely, but rather the use of Artificial Intelligence (AI) in Kenya’s electric vehicle (EV) sector. While the realities between Kenya and China differ, this conversation is to help us understand the level of AI use and what the future looks like for all EV owners and users.
Warren Ondaje, the Managing Director of Africa E-Mobility Alliance (AfEMA), joins us now for this conversation. He also runs the Carbon Leapfrog podcast and Newsletter, which focus on clean transport and energy on the African continent.
Hello and welcome to The Africa EV Show, Warren!
Warren Ondaje: Thanks, Njenga. This is a pleasure. I look forward to this conversation and I’m glad to be here.
Njenga: Before we get started with how AI is being used or how it will be used in cars, particularly in EVs here in Kenya, let’s first take a step back to break down a few key concepts. Now, a lot of us are familiar with AI tools like ChatGPT and even Google Maps now uses AI. But what does it mean in the context of a car?
How is AI being integrated into vehicles in Kenya?
Ondaje: AI sometimes can be interpreted as a buzzword because it’s probably increased in attention following the rapid development of computing. And now we are seeing this also make its way to vehicles. AI simply stands for artificial intelligence.
In vehicles, this has primarily also taken off in the private use vehicles, but EVs have predominantly been the ones that are most integratable with AI, simply because these are all electronic or electrical devices, right from the batteries to the drive system to the in-vehicle experience. So AI for vehicles simply means that this technology allows the vehicle to automatically process data, make decisions and even take some actions either autonomously or semi-autonomously. And goes back to your point, and this can be in multiple levels, right from level zero all the way to level five.
Njenga: We see a lot of cool tech in cars in China, the US and Germany, but not as much here in Kenya where a lot of the vehicles are older and often secondhand. How long will we have to wait for these new AI enhanced vehicles to make it to Kenya and other African countries? AI appears to be a far-fetched idea.
Ondaje: And maybe some people in the audience might think that this technology is far from getting to us in the developing economies. But this might not necessarily be true because we already have EVs on our roads, particularly in the public, shared and commercial spaces with electric buses, motorcycles on our roads, as well as even electric taxis. These are the fastest-growing segments for EV adoption across the continent.
But then these vehicles are also new vehicles and modern vehicles with these capabilities. So you and I might not necessarily experience AI in the vehicle as if it was our personal vehicle, where you’d speak with a voice assistant in the car or maybe get lane assistance. But in these public, shared and commercial vehicles, mainly the four-wheelers, buses and potentially taxis, there are certain levels of automation that are happening in those vehicles to do with safety, to do with vehicle background operations.
So you might find that there are operators that are busy working on crunching data about battery performance, driver behavior, route optimization to ensure that the vehicle is operating at the best possible level to execute those operations. So the AI experience might not be in vehicle. It might probably be external and in the background.
Njenga: What I am hearing, because also these vehicles are the same ones that will be imported into Kenya. I don’t think we are building any new EVs in Kenya anytime soon. But I think AI in terms of like battery systems and optimization, that is what you are alluding to.
We may not see AI that we are interacting with, like take me from point A to point B, but in the like running of the systems within the car, that’s my understanding of what you just said.
Ondaje: Yeah, correct. So the experience might not be at such an individual level, but it’s more of managing the asset that is on the road because someone else is operating this vehicle. It’s a taxi driver, it’s a bus driver, and they’re supported by a larger team that is tech driven behind the scenes, actually looking at those vehicles and optimizing the operations.
Before we see the real experience that probably you and some people in the audience might be alluding to or talking about is the one that you’re experiencing in the car. The service about cruise control, the lane assist, the assistant that is in the car and listens to you and rolls down the window or you point to the screen and they do something with it. That is something that will take place in the private vehicle ownership.
And that is completely market driven. So today there are people who can afford that. I’m sure we’ve seen such high end vehicles on our roads, even in Kenya.
And these are the same people who could immediately access those high end vehicles. But the trend from global development is that this technology is becoming more and more accessible to even mid range or low range vehicle types. Your normal small compact vehicles are now coming with much better features tech wise compared to some of the older ICE models.
So I think sooner than we think we will have these technologies already in our cars because we’re already experiencing some level of low level AI, level two, probably with the assisted cruise control features, and maybe some have some voice assistance. So the technology is coming and it’s coming quickly. But on a larger scale, we’re already experiencing it, especially if you’re using public, shared or commercial transport.
Njenga: Unfortunately, where I come from, you and I, is that the politicians whose lifestyles we pay for are the ones who can afford this vehicle. But anyway, beyond the car itself, are there uses for AI in optimizing EV charging infrastructure and reducing rage anxiety for Kenyan drivers? I know we’ve mentioned battery systems.
Ondaje: This is also not very spoken about or illustrated because a lot of the management and assistance of AI tools is actually in the background by some of the operators on the ground. One of the opportunities is on optimizing charging. Nowadays, these EVs, especially the four wheelers, come with smart charging features that allow them to charge during periods that the power on the grid is most affordable.
As you’re aware, in Kenya, there’s an e-mobility tariff that only starts at a certain time. That’s 10 p.m. to 6 a.m. And that’s the off-peak period where the tariff is actually reduced. Now, one way to implement this is by manually switching on your charger at that point every night and accessing the tariff or simply going to bed and allowing the vehicle to prepare itself and kickstart the charging process once that time is reached.
And it might not exactly be at 10 o’clock. It might even find a more suitable time where your battery is cooled and it’s at an optimum stage. The other opportunity is on predictive maintenance, especially for this is now background function that EVs as they move around, they’re collecting a lot of data around their location, the terrain that they’re experiencing.
You might even capture speed metrics around acceleration, deceleration. There could even be hard cornering. And this data can now be compiled to form, number one, a profile of who is driving this vehicle.
And it can tell you how risky that driver is basically, including if that’s you. But then number two, it can also start throwing up recommendations on ways of maintaining or operating your vehicle so that you’re conserving the battery, for instance, or reducing your exposure to wear and tear because of probably the route that you’re using or the times that you’re driving if they end up affecting the performance of your vehicle. The other goes now to the charging infrastructure, and this might address some of the range anxiety concerns.
From your charging patterns and your driving habits, AI can also use this data in the background to start modeling, and this is now for the charging operators, potential positions for actually locating your battery, swap stations, or your charging points. This is important because nowadays we have Google Maps, and that is a great tool for route optimization. Sometimes it even shows you where traffic congestion is happening.
And this existing infrastructure can still be layered on top with information around EV operations to arrive at which are the best locations to actually locate this charging station so that people are not running out of charging traffic or driving longer than expected distances without necessarily planning or anticipating an EV. The last one is, again, as electric mobility is now taking shape, data from the vehicle can finally be combined with the user behavior to come up with something like a driving profile. Because after understanding the vehicle and how it’s being operated, together with some technology even to do with visual inspection of the driver behavior, then you can have a comprehensive overview of what is user AI like when they drive this vehicle.
And this information can be useful, especially for those who are financing vehicles to understand the risk profile of the drivers, or those who are employing drivers to drive their vehicles for different purposes.
Njenga: I think from what you’ve just said that insurers will be ripping big from this AI thing, right? Potentially.
Ondaje: It’s still a technological opportunity, given the fact that initially with internal combustion engine vehicles, the most insurers could see, at least those that were tech driven, was GPS data. This was just showing you the speeds and where this vehicle is located. But with EVs, you’re going a step further to now look at the energy dynamics in this vehicle.
You could not previously see how much fuel is this guy consuming, when are they fueling, but now you can. And you can even, if you’re very interested, see where this energy is coming from. Is it coming from renewable energy sources or non-renewable?
So combining this end-to-end understanding of the energy sources and the energy charging behavior, together with the driving behavior, can help make insurers smarter about how they issue their premiums, who they issue their premiums, and how to better incentivize those who are good users of the asset and those who are poorer asset users can be penalized.
Njenga: Let’s get back to the cars and Chinese cars seem to offer a lot of tech for the least amount of money. I wish I was in China this time. And when you talk about AI enhanced cars coming to Kenya and other African countries, is this mostly going to come from China or should we expect the legacy players like Japan, Korea and Europe to also be active in this market?
I know, of course, there are vehicles that we have currently that have cruise control, lane assist and things like that. But do you think that this AI shift will majorly be Chinese EVs or everyone across the board?
Ondaje: Certainly, as you can see, YouTube can be the best exposer of some of these technologies. And every so often you will see a Chinese OEM showcasing the capabilities of their vehicles. And by far, China now has the largest number of brands that are EV based.
So the competition is definitely stiff from the Chinese market, meaning that they are constantly pushing the edge. We’ve seen this recently with the BYD five minutes charge, putting in 400 kilometers worth of range. And this is already triggering a change in even perception about what’s the future of charging going to be like.
Do we still need to be relying on huge batteries or just faster speeds of charging to bridge this gap of range anxiety? So China is definitely one of the fastest moving markets. And China has many advantages, starting from control of the supply chains.
And now the innovation that’s happening with the sheer number of companies coming out of that part of the world. At the same time, Europe, I was just at the Nordic EV Summit about 10 days ago. And the developments from the European market are also moving quite quickly.
Driven by the bigger brands, BMW, Stellantis, VW. They’re all exploring, number one, capitalizing on the commitments that the European market has in transitioning to full EVs and banning progressively the use of ICE vehicles from 2035. And this is going to be a big moment when the European markets actually shift because they’ve driven a large part of automotive demand globally.
So the European market is certainly moving. It’s not catching up as quickly to China because even China is proving to be a fast mover to the developing world. Latin America, now Africa, there’s more partnerships for setting up EV facilities, charging infrastructure and vehicles compared to some of the European brands that are there.
Let’s also not forget Asia, other parts of Asia. Korea is moving quite quickly with Hyundai, Kia brands. These are producing, and their models are also fantastic.
But Japan, that’s a legacy operator with Toyota, Nissan, Honda that have not moved as quickly. And it’s already, their lunch is basically eaten by someone else. And this is going to change the future of automotive dynamics on the continent and in the world really because we are getting to this new age of new vehicle technologies that are coming up from newer parts of the world.
And this is also changing how collaborations and maybe even procurement and even supply chains will now operate as people shift away from buying used vehicles from Japan traditionally and shifting now to these new emerging markets from the East and potentially from Europe.
Njenga: Do you expect that autonomous driving will ever become a thing in Kenya? It seems hard to imagine given how much infrastructure is needed to make it safe, which they’ve done in places like China. But you are still very far behind.
Ondaje: Autonomous driving for the African continent, let alone the Kenyan context, is still faced with its own barriers. We should first appreciate how much investment it takes to just have an allowing infrastructure. You need to have smart roads.
And these are roads that should not have the basic flaws that ours have, where potholes, markings on the lanes, lights that are actually working, together with excellent mapping and clear knowledge of the road system. Because in Africa and Kenya, we have a mix of both tarmacked and non-tarmacked or non-paved roads. These are fundamental ingredients that we are missing.
The final one is also access to internet. Smart roads, smart connectivity requires robust, affordable and high-speed internet. Something that we’re still grappling with as a country, as a continent.
So these are significant barriers, but it also doesn’t mean that it’s going to be a complete no-no. I think the opportunity here could start by, similar to how it started out in China, the technology can be implemented in closed environment or controlled environment. So large campuses, this could either be in learning institutions, it could be in ports or other industrial facilities that can give a small test bed to just see how well can this infrastructure actually operate.
The other is, for now, we can only lean on the lower levels of AI, level two and three, adaptive cruise control, lane assist, automated parking. These are potential low, high influence that some vehicles already have. So we can have a mixture of graduating from the low levels that can work even with tougher terrain, but also testing at the high level.
It’s only that it also needs some investment commitments, interest and a way of proving sustainability because we can’t just implement because it’s nice to implement level five. It actually needs to make some commercial sense to the end users or to the operators.
Njenga: With so much tech now in cars, there is growing consumer concern about what happens to all the data because so much data is being collected. This is produced by the vehicles or when driving these vehicles, or about your habits as a driver or consumption habits in terms of how much you charge and how soon you run out of that charge and things like that. And we know that in countries in Europe and China, there are a lot of restrictions on what companies and governments can do with that data.
In Kenya, we have a data privacy entity that is government, but what else is there about data privacy in Kenya, especially when it comes to EVs?
Ondaje: So far, Kenya has actually been quite progressive with data protection, and this is something that is to be emulated across the African continent. We have the Data Protection Act of 2019, which from the initial framework of developing further regulations that will now address privacy, registration of data controllers, processors, and these frameworks came to happen. So we have the data protection regulations focused on registration of data controllers and processors, complaints and handling.
Now there’s one that’s in discussion about compliance or conducting compliance audits. The Act also initiated the Office of the Data Protection Commissioner or DPC that was in charge of now handling all the regulations and compliance measures. So that provides quite a robust framework because it was triggered from other sectors that preceded the EV sector, telecoms, any other internet facilities, even financial sectors, because these are still large processors and consumers of personalized data for offering their services.
I think with the EV market, there shouldn’t be cause for concern because there’s already existing frameworks. I think it’s a matter of now going to review from a policy perspective where the gaps as this EV adoption is coming up. Do we have any specific clauses that speak to the automotive sector?
Or do we have any that talk to this transport? Because with AI, with autonomous vehicles, connected vehicles, we’re definitely going to be collecting lots of data, especially the companies that are selling these vehicles. And there could be third parties that are interested in this data.
So it’s important to actually work along these frameworks. But that should not be really cause for concern at the moment because we also have such a small fleet and a low level of adoption. I think it’s important to start addressing, but also understand that there are frameworks that we can count on as a country that can cover the consumer.
Njenga: So I buy this car that has all these AI features in it, and then it breaks down or something goes wrong in Kenya or in any African country. I know that South Africa is a bit different. You could get many of these solutions or services.
But in Kenya and many other African countries, do we have enough skilled mechanics who can handle these AI enhanced vehicles? And as a customer, that’s something that I would be very worried about.
Ondaje: So the future automotive repair maintenance for the EV space is shifting away from nuts, bolts and hammers, but it’s now incorporating more laptops, phones and the internet, particularly because all these devices are connected and have a smart function that needs to be maintained, updated, debugged every so often. The good thing with electric vehicles is that already the technology itself is quite simple. You have a motor, battery, a controller, and all this just works together to drive the vehicle.
Compared to an ICE vehicle that has thousands of individual parts that need to be put together and coordinated to burn a liquid, turn it to heat, find ways of turning that heat and the chemical reaction into motion. It’s a more complicated process. So with EVs with fewer moving parts and a less complicated infrastructure, there’s less maintenance that is inherent with the vehicle.
There’s also a cost reduction because of the amount of energy it takes. It’s more efficient. Therefore, it will overall in its lifetime use less energy compared to an ICE vehicle.
When it comes to the connectedness nature and the tech tools, including artificial intelligence that are built into these vehicles, it is partly the responsibility of the customer together with the seller of this vehicle, because the seller of the vehicle is also responsible for maintaining the integrity of the software, of the sets of software that run this vehicle. So it’s on the company first to inform the user about the data that’s been collected, get their consent, and maintain an open channel of communication on where that data is being used and what sort of data is being used from the user. If that is not done correctly, then it can lead to issues to do with data protection and privacy.
But also this data is important for some of the technicians or the mechanics that are working on these vehicles because they need to understand, for instance, what are the errors that are popping up and how can they be resolved. Today, because of the relatively well-developed ecosystem globally, it’s not impossible to find the capacity in Kenya or in Africa is sufficient to service these EVs. Because it’s a connected vehicle, you also don’t necessarily rely on people that are in the country to fix the vehicle.
You just need a vehicle that its software components can be connected or understood remotely. The hardware parts, the moving parts, those can still be serviced by existing mechanics. So the question that adds on top of whether we have an existing ecosystem to maintain the vehicles is also whether we have enough partnerships that allow the importers or the current dealers of EVs to collaborate with the existing ecosystem that can maintain the usual tires, steering wheel, the usual mechanical parts, and then build on top a team that can focus on the more software-based components.
So it’s possible and I believe it’s sufficient. It is also market-driven. The more vehicles, the more the capacity comes up.
Njenga: So nobody is losing their jobs yet. We can be assured of that.
Ondaje: Not yet. Funny thing is we might probably end up with a situation where there’s less maintenance jobs that are needed. This has been cause for fear for some dealers who are reliant on the servicing fee and maintenance fee as part of their service contracts.
So with this change, it’s important to now identify what other services can replace the reliance on the traditional maintenance and repairs. So this could be in providing other digital services or just retooling as an organization to find other opportunities because it’s a vehicle that’s inherently more efficient. So it’s just not going to need as much support to keep running because it will run for a good amount of time.
Njenga: Thank you so much, Warren, for this conversation. I know it’s been a long time coming, almost a month and some few days, but here we are and we will keep having this conversation.
But guys, some of these conversations can go on and on, but it’s a wrap now on this discussion about AI’s transformative or potential transformative role in Kenya’s EV sector.
Outro: From optimizing battery life in electric buses to AI-driven fleet management, revolutionizing Matatu routes, we’ve seen how technology is rewriting the rules of mobility. However, we must remember that the road ahead requires more than just algorithms. It requires collaboration, policy, and local innovation.
Here is the big thing. Data, AI isn’t just a buzzword in Kenya’s EV boom. It should be a tool for equity by ensuring that no one is locked out, whether it’s startups like Roam using AI to predict battery swaps or Kenyan researchers developing homegrown solutions.
AI should make all this possible without discrimination. But we also must ask, who controls the data? How do we train our mechanics for AI-powered EVs?
And can Kenya leverage partnerships with China and the Western countries without losing sovereignty over its tech future? A huge thanks to our guests today for sparking these critical conversations. To our listeners, keep pushing these questions, demand policies that protect data, invest in local talent, and ensure AI serves Kenyan needs, not just global corporate interests.
If you’ve loved this episode, share it with a friend, leave us a review, and tag us on social media. Next week, we host another guest who will be sharing on the African EV space and giving us another perspective on another topic that we’ll be looking at. Before then, head over to our website www.chinaglobalsouth.com for more of our content, which covers excellent research on China in the Global South, tech, and many other interesting topics by our team.
Also, be sure to shop around our YouTube channel for more interviews and insightful stories on geopolitics, technological advances, and even critical resources like lithium and cobalt, which are essential for powering electric vehicles. Thank you for being with us on the Africa EV Show.
My name is Njenga Hakenah, and I hope to see you again soon.
