Lalrinngheti Sangsiama
In June 2025 I went to Bali to make a vehicle.
Or — more honestly — I went to Kul Kul Farm, in Ubud, to learn how to read a culm, and the vehicle was the brief I gave myself to make the reading concrete. The Bamboo U eleven-day course is structured as four days of foundation — forestry, treatment, scale modelling, joinery — followed by six days of full-scale building with the resident master carpenters. On the build days each student chooses a project. Most build a shelter, a chair, a small pavilion. I had been carrying around a tawlailir for years, and the course was the first place I could put it in physical material.
The tawlailir is the Mizo single-axle cart. A wooden bed, two wooden wheels, a long shaft, drawn by hand. Everyone in the village has it or knows where to borrow one — and it is, for that reason, a more honest piece of material culture than most of the things called heritage. Farmers in central Mizoram routinely cover up to four hours a day on foot between farm, home, and market. The tawlailir is what they push, drag, or load when they need to move more than their own bodies can carry. I had been wanting to redraw it, in bamboo, for the present, and to make the redrawing strict enough that the design decisions could not hide behind metal.
Before Bali I had built one steel-framed prototype with bamboo elements on top of it. It rolled. It carried load. It also disappointed me, because the steel was doing the structural thinking and the bamboo was performing it. I wanted a build where the bamboo was load-bearing in every joint, in every wheel, in every fastener. The course gave me the species range and the carpentry to try.
the brief
Make the whole vehicle in bamboo. Frame, wheels, joinery, fasteners — all of it. No metal. Carry useful load. Take a battery and a motor. Climb a hill.
That last requirement matters because central Mizoram is not flat. A mobility object that only works on level ground is, for our context, a mobility object that does not work. Putting a battery and a small motor into the design meant the prototype had to survive vibration and torque the tawlailir was never asked to handle, and that the joinery had to hold against forces well beyond what a hand-pulled cart would generate.
The brief was strict on purpose. A redrawing that leans on bolts and brackets is an honest piece of design, but it tells you nothing new about bamboo. A redrawing that refuses metal tells you, very quickly, where the species you are working with does and does not earn its place. The brief was a way of forcing the culm to argue for itself.
the five species
Bamboo U sits inside Indonesia’s bamboo economy, which is an order of magnitude more developed than the Northeast-Indian one. Five species from the campus stock went into the scooter, each pulling its own weight.
Dendrocalamus asper (bambu petung) — the structural workhorse. Thick walls, large diameter, dense fibre at the outer skin. The main frame members were drawn from this. It is the closest analogue in the Indonesian range to the load-bearing role Dendrocalamus hamiltonii plays in our own forests.
Dendrocalamus asper niger (bambu petung hitam) — the same species, in its black-skinned cultivar. Used where the frame met itself in visible joinery, partly for the colour, mostly because the older clumps the black culms came from gave straighter, more disciplined sections.
Gigantochloa apus (bambu tali) — long, flexible, fine-walled. The lashings and the bindings were cut from this; the name tali in Bahasa means “rope”, and the species has earned it. G. apus splits with the grain in long predictable strips that hold under tension without snapping.
Gigantochloa atroviolacea (bambu tali hitam) — the same role, again in the dark cultivar, used where I wanted the binding visible and contrasting against the petung.
Thyrsostachys siamensis (bambu jakarta) — thinner-walled, lighter, with a finer node spacing. Used for the secondary structure and the wheel laminations, where weight mattered more than absolute strength.
Five species, five jobs. The work the course did, before any tool came out, was to teach me that a vehicle in bamboo is not a vehicle in one bamboo. It is a vehicle in a small assembly of species, each chosen for the load and the geometry it is going to inhabit. That is the part of the discipline you do not get from a textbook. You get it from a master carpenter pointing at one culm and saying this one, here and pointing at another and saying not that one, never there.
The carpenters who walked me through this were Asmara Jaya and Moko Sumerta, with Mike Linthon and Luis Echeverría on the design side. They are the reason the build held.
the joinery
No metal meant the joints had to be drawn before the cuts were made. Three families did the work.
Whole joints and fishmouths. Where one petung met another at an angle, the meeting culm was cut to a fishmouth that wrapped the receiving culm’s curvature and bedded against the node. The geometry of the cut, more than any fastener, is what carries the load — the wood-joiner’s principle that the joint is the wood, not the thing holding the wood together.
Lashings. Where members crossed, G. apus strips were wrapped and tensioned against each other, the way Cheraw poles are held against each other in the dance, except here the binding has to hold under driving torque and not just rhythmic clap. The course teaches a specific lashing geometry — figure-eight wraps, locked at intervals — that uses the friction of the bamboo skin against itself as the principal mechanism.
Bamboo pins. Where a fishmouth alone would not lock against rotation, a bamboo pin driven through pre-drilled holes carried the shear. The pins were cut from older, denser petung. A bamboo pin in a bamboo socket is not a metal bolt and should not be designed as one — its tolerance, its expansion behaviour, and its failure mode are different. The carpenters teach this almost as a separate craft.
The wheels were laminated. Strips of T. siamensis steamed, bent, and glued into a hoop, the hoop trued against an axle stub turned from petung. Two wheels, no spokes, the laminations forming the structural ring and the tread the same surface.
what it carried
The evidence the scooter produced is that the geometry holds, the joinery holds, and the species selection does what I asked it to do.
It answered: can the tawlailir geometry be drawn in bamboo, end to end, without metal? Yes, under the loads tested, in the species available at the Kul Kul Farm stock.
It did not answer: will the same design hold for years in Mizo monsoon, against borers we have and Indonesia does not, on roads steeper and rougher than the campus tracks? That is a different question, and the bamboo it would have to be drawn in is not D. asper but Melocanna baccifera and D. hamiltonii, with different mechanical envelopes and different treatment needs. The Bali scooter is the geometric proof. The Mizo scooter is a separate piece of engineering, and the work on it has its own life now in the bambam kart, in build with NOA Mobility, where the species, the loads, and the geometry are all being re-derived for our own context.
what the scooter answered for the studio
Three things, in order of how much they have shaped the work since.
The first is that a serious bamboo build does not start with a single species. It starts with a small palette, chosen for the loads and the geometries the design needs, and the palette must be assembled before the joinery can be drawn. The traditional Mizo builders knew this. The Bamboo U carpenters know this. Most of the contemporary bamboo discourse, particularly the architectural-magazine version of it, hides this — partly because the photogenic culm is usually a single species and partly because the studios producing the photographs are working in regions with only one or two viable structural bamboos. In Mizoram we have at least seven species in the studio’s working set, with different fibre orientations, different node spacings, and different ages-at-harvest. The Bali scooter taught me that the discipline of species selection is not a refinement layer on top of design. It is the first layer.
The second is that the no-metal constraint is a teacher. A design that is forced to refuse metal asks more of the bamboo than it would otherwise have to give, and the asking is what produces understanding. I do not believe the tawlailir on Mizo roads will be a no-metal object — the bambam kart will use carefully selected steel where load and safety demand it — but the discipline of having drawn one all-bamboo prototype changes how the studio thinks about the hybrid versions that come after. You know, then, what each material is doing in the assembly, and what each is being asked to do for reasons of convenience or cost rather than function. The metal that survives the discipline is the metal that earns its place.
The third is that the tawlailir is a more capacious object than the romantic version of it lets you see. It is not just a cart. It is a chassis logic — a single-axle, low-slung, hand-or-power-pulled platform for moving load between fields and home — and the chassis logic can be redrawn for different ends. A scooter is one version. A cart with a bed and high sides is another. A small mobility object for a farmer who is also a vendor at the Aizawl bara bazar is a third. The geometric question, once answered, opens the design space; the bambam kart is one occupation of that space, but it is not the only one.
what comes next
The Bali scooter sits at Bamboo U as a course piece. It will probably end its working life there, because moving a vehicle of that mass across borders to a different climate and a different bamboo economy is not the point of it. The point of it was the answer it produced for the studio, and that answer is now travelling in the form of the bambam kart and the wider mobility line the studio is pulling out of the tawlailir.
What I take from Bali, beyond the prototype, is a working knowledge of five species I did not previously know in my hands; a vocabulary of joinery that maps onto the Mizo joinery I grew up around but extends it; and a clearer sense of which questions the geometry alone can answer and which questions are waiting for the engineering. The course did what an eleven-day course can do at its best. It did not give me a vehicle. It gave me the discipline of how to draw one.
The work continues. The next cart will be heavier, slower, less photogenic, and more useful. That is the trade.
Companion essays: the regeneration model, bamboo is not timber. Related work: bambam kart, bamboo u, bali.