Insight, E-Mobility
An open battery-swapping network in Kenya solves the fragmentation problem that has been holding African motorcycle electrification back. Open architecture is a technical property. Whether the standard is governed is a separate question, and it is still open.
SUN Mobility and Vivo Energy have launched an open-architecture battery-swapping network in Kenya. Company announcements at launch describe roughly 35 stations operating across Nairobi and Mombasa, vehicles from more than ten manufacturers able to use the system, and around twenty of the sites at existing Shell forecourts. Vivo Energy itself operates Shell- and Engen-branded sites across more than 20 African countries, which is the distribution asset behind the partnership.
The station count is the headline. The manufacturer count is the story.
African motorcycle electrification has been heading towards a predictable and expensive failure mode, in which every vehicle manufacturer builds its own proprietary battery and its own swap network. Each firm then pays the full cost of infrastructure while serving a fraction of the riders, every station runs at low utilisation, and the rider is locked to whichever brand happens to have a station near their route. Nobody reaches scale, the capital requirement per vehicle sold stays absurd, and the sector stalls for reasons that have nothing to do with the technology.
A shared standard breaks all three constraints at once. A manufacturer can sell a motorcycle without raising the capital to build a swap network. A station can serve the whole fleet in its catchment rather than one brand's share of it, which is the difference between a stranded asset and a functioning one. A rider can buy on the merits of the vehicle. This is a genuinely important development and considerably more consequential than another electric-motorcycle launch.
Which is exactly why the governance question deserves asking now, while the network is 35 stations rather than 3,000.
Open architecture describes a technical property: the interface is documented, and parties other than the owner can build to it. That is real and useful.
It says nothing about who decides what the interface is, how it changes, who is admitted, on what terms, and what recourse anybody has when the answer is unwelcome. Those are governance questions, and they are separate from the technical ones in a way that repeatedly surprises people who assume that publishing a specification is the same as sharing control of it.
The history of technical standards is largely a history of this distinction being learned the hard way. A specification controlled by a single firm is open in the sense that others may use it, and closed in the sense that they cannot influence it. Both statements are true simultaneously, and which one matters depends entirely on whether the standard-holder's interests stay aligned with yours.
Shared infrastructure that many parties depend on, that none of them individually owns, and that degrades if it is mismanaged is not a new problem. Elinor Ostrom spent a career documenting how communities govern exactly this class of resource successfully, and the conditions she identified are specific rather than vague.
Clearly defined boundaries around who may use the resource. Collective-choice arrangements in which those affected by the rules participate in making them. Monitoring that the users trust. Graduated sanctions rather than arbitrary exclusion. Accessible and low-cost conflict resolution. And recognition by outside authorities of the users' right to organise their own governance.
Set the current arrangement against that list honestly. A shared battery-swapping network in Kenya has, at present, a bilateral commercial contract between two firms. Manufacturers are customers of the specification rather than participants in setting it. There is no published change procedure, no notice period, no independent monitoring, no conflict-resolution mechanism that does not run through one of the two counterparties, and no recognised body through which the manufacturers using the standard could organise a collective position.
None of that is a failing on anybody's part. It is simply what a young commercial partnership looks like before anyone has decided that it is infrastructure. The point is that it will become infrastructure, quite quickly, and the moment to install governance is before the dependency is large enough to make governance politically contested.
The two layers where surplus accumulates are the interface specification and the physical site. Neither is a battery. Manufacturers building to a shared interface are making a sunk commitment at the commodity end of the stack, in favour of parties at the other end.
A manufacturer that designs a motorcycle around a swap interface has made a commitment it cannot cheaply reverse. Battery-bay geometry, connector, thermal design, mounting, control electronics, and communications protocol all follow from the specification. Redesigning around a different interface means retooling and requalifying a vehicle that is already in the market, and stranding every unit already sold.
The standard-holder has made no equivalent commitment. Its stations can be retrofitted, its specification can be revised, and its commercial terms can be adjusted at renewal.
That asymmetry is not an accusation. It is a description of where the option value sits, and it holds regardless of anybody's intentions or good faith. Good faith is not a governance mechanism, because it does not survive a change of ownership, a change of strategy, or a change of the people who made the original arrangement. Structures survive those things. Intentions do not.
This is the same shape of problem we described in platform-anchored charging networks, where the party that solves a real coordination failure also acquires a durable position from having solved it. The response there and here is identical, and it is not suspicion. It is that arrangements of this kind should be written down properly while everybody still likes each other.
There is a second thing in this announcement that deserves more attention than the swap technology.
A swap station needs an urban plot in a high-traffic location, a grid connection with meaningful capacity, physical security for valuable inventory, staffing for long hours, licensing, and a location riders already know how to find. In Nairobi and Mombasa, and in most African cities, the set of sites meeting all of those conditions is small and already occupied.
Fuel forecourts meet every one of them. They have the land, the power connection, the security infrastructure, the staff, the permits and, importantly, the habit. A rider already knows where the Shell station is.
That is why the Vivo Energy footprint matters more than the swap technology. It is a distribution asset that no e-mobility company could assemble in a decade at any plausible cost of capital, and it means the practical gatekeeper of motorcycle electrification across much of the continent may turn out to be an incumbent fuel retailer.
This is not a bad outcome. Incumbent assets being redeployed into a transition is usually how transitions actually happen at speed, and we have argued that what travels with an incumbent is worth examining precisely because the redeployment is valuable enough to be worth getting right. But it does change what the strategic question is. Manufacturers negotiating access to a battery standard are, in substance, negotiating access to a property portfolio.
The useful thing about this problem is that it is well understood in other sectors and the remedies are unexciting. Four of them would cover most of it.
A specification body with manufacturer representation. Not a users' forum with advisory status. A body with a defined membership, a published procedure for proposing and approving changes to the interface, and voting rights that mean something. Manufacturers who have sunk tooling into the standard have a legitimate claim to a vote on its future.
Notice periods and backwards-compatibility commitments. A specification change that obsoletes vehicles already in the field imposes a cost on manufacturers and riders that the standard-holder does not bear. A binding notice period, and a commitment that existing vehicles remain serviceable for a defined term, converts an open-ended exposure into a manageable one.
Published admission criteria and non-discriminatory pricing. Which manufacturers may join, on what technical basis, at what cost, decided against published criteria rather than case by case. The value of an open network to a new entrant depends entirely on whether entry is a right or a favour.
Continuity arrangements. What happens to the specification if either firm is acquired, exits the market, or changes strategy. An escrow of the specification, or a licence that survives a change of control, is cheap to arrange now and impossible to arrange later.
None of this is hostile to the companies involved. A governed standard is more valuable to SUN Mobility and Vivo than an ungoverned one, because manufacturers will commit more capital to an interface whose future they can see, and the whole proposition depends on manufacturers committing capital.
Kenya is the right place for this to be settled, and the moment is unusually favourable.
The country has a large and growing commercial motorcycle fleet, an active regulator, a manufacturing base, and a set of operators with real field experience. It also has, at 35 stations, a network small enough that governance can still be designed rather than retrofitted. Once several hundred thousand vehicles have been built to a specification, the governance of that specification becomes a matter of public interest whether or not anybody planned it that way, and it will be settled by whoever has the most leverage at the time rather than by whoever has the best argument now.
Standards are the cheapest industrial policy available to a government, and the window in which they can be shaped is short. The riders whose daily earnings will depend on this network for the next twenty years are not represented in any room where the specification is currently discussed, and their interests, which run to price, availability, and the freedom to change brand, are the ones most likely to be traded away quietly. Our fieldwork with commercial riders in Nairobi starts from what that daily calculation actually looks like, and it is the perspective least present in standards negotiations everywhere.
The Lab works on this in e-mobility and transport.
If you are a manufacturer, a regulator, or a funder with a position in this network, the questions above are worth putting in writing now, while the answers are still cheap.
This is an independent insight piece by Transitions Lab. For the Lab's applied work, see E-Mobility & Transport. See also Own the Battery, Rent the Shopfront on the make-or-buy decision inside a swap network, and The Ban Is Not the Policy on the same monopsony question arriving in mineral processing. To discuss a study, see Contact.