Category Archives: Technology


Barely a month ago, I started working on the idea of a pedal-powered tractor for rice transplanting. This was for my Live-in-Lab project in India. I have just spent four days in the manufacturing workshop in a campus in Tamil Nadu, a 10h trip from my main campus. This post is tells how the manufacturing went on.

Manufacturing overview document and CAD drawings...
Project presentation and blueprints for manufacturing…

Day 1

After a short month of design work on my project, we are already starting manufacture! Aymeric, who is also doing a project in my lab, and who has already started manufacture announces in a stern voice: “one centimetre”. That is the precision with which our projects are being created. I’m not sure whether I should laugh or cry. As a microengineering student, I feel kind of sad.

In the manufacturing workshop, as well as the “Indian precision”, I also notice the Indian safety standards, or rather the lack of any. Here, the workers are bare feet or wearing sandals, and none of them is wearing safety gloves or glasses. The noise of the machines is deafening, and I often find myself plugging my ears while averting my gaze to avoid the bright flashes from the welding. I would be happy if I don’t become deaf and blind by the end of this project!

The Indians don’t have workbenches: they work directly on the ground. To discuss the design, we crouch besides the parts we’re talking about.

Discussions à propos d'une pièce. accroupis dans l'atelier de manufacture.
Discussing the design of a part, crouching on the floor.

My project is a rice planting machine. When I started doing my research about what already exists, I was disappointed to discover many machines already existed. From cheap machines to massive agricultural monsters, the subject seemed under control. Additionally, while checking out Youtube, I found so many student projects it seemed like all Indian MTech students have to build a rice planting machine. What is the purpose of my project then? I couldn’t see how my work could serve the subject. So I decided to take a radically different approach. Rather than building an average machine that farmers would push or pull across the paddies, I decided to create a pedal-powered machine. Based on a bicycle, with a transplanting trailer, I am convinced that my design could do the job. With countless hours of pedalling in mud as a mountain biking competitor, I knew it was possible to ride through a thick layer of mud. Now I had to prove them.

On rencontre l'artisan/ouvrier qui va construire la majeure partie de mon vélo...
First meeting with my project’s maufacturer…

Apparently, manufacture starts with a prayer. I join my palms in front of my chest, close my eyes, and listen to the prayer in Malayalam. I even recognise some sentences after a month and a half in the Ashram!

We start by building the back wheels. Soon enough, the long strips of steel get bent and welded together as tractor wheels. Yesterday, we went to get a second-hand bike. I like its “old school” looks, but it will soon be disassembled by a gang on young Indian workers.  Only the main parts are kept for my design.

Mon nouveau vélo... qui sera vite démonté !
My new bike… while it was still in one piece!
Voilà comment on fabrique des roues ! Des grandes bandes de métal sont courbées à la main, puis soudées aux rayons et au centre de la roue.
This is how wheels are made! By bending strips of steel then welding everything together..
Construction des roues. La précision : pas plus que la largeur du trait à la craie.
Building the wheels. Precision? Not more thant the width of a chalk mark.
Tout feu tout flammes : qui ne voudrait pas être ingénieur quand son équipe de manufacture fait cracher du feu à un simple bicyclette ?
Who wouldn’t like to be an engineer when your manufacturing team can turn your bike into a fire-spitting dragon?

Day 2

The manufacturing is going well, and my Ricycle is starting to take shape! When I arrived in the morning, I discovered with awe that my back wheels were almost finished, and they are impressive. But when I try to lift one up, suddenly, it isn’t as stylish: they are almost as heavy as my travel backpack. At 13kg each (ok, a bit less than y backpack), I start to lose faith in my project… How will my small farmers manage to pedal this prototype through the fields?! A sturdy mountain biking wheel weighs less than 2kg, I would never have imagined mine weighing more than double this figure! My supervisor tries to comfort me: “worst case, we give your system to an ox, it will be able to pull it!”. Cows are sacred in India, the ox gets all the work. I still haven’t understood the difference, but if they say so…

En arrivant les deuxièmejour, mes roues sont presque terminées ! A l'atelier, il n'y a guère d'équipement de sécurité, et mes yeux souffrent à chaque fois qu'ils s'arrêtent par mégarde sur un poste de soudure. Comme il y en a partout, c'est très dérangeant. Je ne suis là que pour quelques jours, je me demande comment les ouvriers font pour ne pas encore être aveugles !
When I arrived on the second morning, my wheels were almost finished! In the workshop, there is almost no safety equipment, and my eyes suffer every time they fall upon a welding station… As they are everywhere, it is very uncomfortable. I am only here for a few days, and I’m wondering whether the workers are turning blind!

The explanation for this general overweight is the materials used for prototyping. While mountain biking wheels are made out of lightweight aluminium, mine are in made out of heavy mild steel. I’ll have to write down the exact name of it. If we use less material, they would not be solid enough… So they’re heavy. Very heavy.

The other materials and parts used for my Ricycle are mostly second-hand and scrap parts from whatever lies around the workshop. For the gears, we swap the front chainrings and back sprockets. The bike will be easier to pedal with this configuration. For the transplanting system, we find some scrap sawing bands and cut them out. I was imagining having flexible blades, but that was in my ideal designer mind. When I see them hammering the strips to flatten them, I understand they won’t be flexible any more. We learnt in first year’s material class that if you work the metal it won’t be as flexible. Oh well…

Contrairement à l'atelier de mécanique où j'ai fait mon stage d'usinage en deuxième année, ici, tout semble être rouillé. Les matériaux sont les moins chers possibles, ce qui change la donne au niveau du poids, de la résistance, etc.
Unlike in the workshop where I did my metalwork course in second year, here, everything seems to be rusty. The materials are cheap, which changes everything regarding weight, strength, etc.
J'avais fièrement décidé d'introduire des lames flexibles dans mon mécanisme afin d'éviter l'utilisation de ressorts (spéciale dédicasse au Professeur Henein). Je trouvais ça très intelligent jusqu'à que je voie comment ils tapaient sur des bandes à scier le bois pour les rendre droites avant de les retordre et espèrer qu'elles soient toujours aussi flexibles...
I was proud to use flexible blades in my design to avoid using moving parts and springs (special thoughts to Professor Henein!). I thought it was very intelligent… until they started hammering the blades in shape, believing it wouldn’t change the material’s flexibility!

Day 3

My prototype is looking more and more like my computer designs, and I’m impressed. Until now, the biggest project I had made was my robot for the Robopoly contest, which had to fit into a cylinder of 30cm in diameter. The Ricycle isn’t anything alike! The bike is elevated by a dozen of centimetres and the back wheels are one metre apart.

Certaines pièces sont très semblables aux dessins que j'ai fournis, d'autres n'ont absolument rien à voir. Tous ont un point commun : un surpoids manifeste qu'en tant que microtechnicienne, je n'avais pas imaginé !
Some parts look very much like the ones I’ve drawn on my computer, other seem to be nothing alike. All have one thing in common: a considerable overweight that as a microengineer, I would never have imagined!

As the day goes on, the project is getting fine-tuned. At first, we set the parts in the right disposition, then we partly weld them (careful with the eyes!), and afterwards the position is adjusted according to additional parts. Everything is getting together thanks to the manufacturer’s work and it is an awesome thing to see. When I have a close look, I notice that it is far from being precision work, and when “fixed” parts are more that one centimetre lose, my microengineer soul dies a bit but then I take a step backwards and am still quite proud of my project.

Ca commence à ressembler à mes dessins 3D sur l'ordinateur !
It is beginning to look like my 3D drawings!
La mise en place des éléments est aussi soumise à la précision
The disposition of parts is also subjected to “Indian precision”! As one of my metalwork teachers would say: “there’s a shitload of sideshift!”.

I feel like a site supervisor, being a snob hidden behind my sunglasses. I have decided to cut myself out by wearing earplugs. My ears thank me for this but the already limited communication is further degraded. The workshop is so noisy that I hurt my ear. And the welding flashes are still as aggressive on the eyes. As I start coughing, I am thinking that with all these metal dust particles in the air I might replenish my iron deficiencies!

During the day, many Indians come to have a look at my tricycle. They try to spin a wheel, to understand how it works… They smile and nod in the typical Indian fashion, and my supervisor explains them what it is. In Tamil. I am almost never included in these conversations, or at the most with a slight nod towards me.

Les indiens sont curieux de mon projet et semblent heureux de le voir prendre forme.
The Indians are curious about my project, and seem to be happy to see it take shape.

Despite the language barrier, I feel some complicity with the other women in the workshop. They are five, two of them are working on machines. We curiously look at each other, and try a shy smile while nodding our heads. Yesterday, at the tea break, they asked me if I spoke Tamil. Unfortunately, I haven’t tried learning. At the chai break, they show me where to get my cup and make a sign to tell me to sit besides them. One of them speaks a bit of English and asks me some questions. Where I come from and for how long am I here. Even if we don’t speak I appreciate this outreach which makes me feel more accepted than in any of the technical conversations about my project in which I was barely addressed.

Day 4

The project is being brought into life! With pride, I get on my tricycle and start riding… before derailing after a few metres.

Une prière s'impose avant de tester le Ricycle pour la première fois.
Of course, we pray before testing the Ricycle for the first time.

The Ricycle has made its first steps! Or rather, its first wheel rotations. I can’t even imagine the efficiency of the machine, every part is lose. We spend the day assembling and disassembling the wheels and axle to align every part and tighten everything up. Little by little, the parts are welded and lined up with washers, and the wheels don’t have a path difference of 10cm when turned in in opposite directions.

My supervisor brought the axle to the lathe workshop to thread its extremities, and the wheels can now be tightened. This also allows the bike to turn, as one wheel slips while the other is rolling forward when the handlebar is turned. Now the bike is not limited to straight lines! Well this might not be a sufficient permanent solution, as when the bike encounters resistance both wheels slip and the bike gets stuck.

Une étudiante indienne a essayé le Ricycle, et ça a l'air de fonctionner aussi pour elle ! Dans ma conception, j'ai voulu prendre en compte la morphologie des fermières indiennes, qui font en moyenne 1m50, afin que le vélo soit adapté à leur taille.
An Indian student tried the Ricycle, and it seems to work for her too! In my design, I wanted to take account of the Indian female morphology, who are around 1m50 heigh.

Besides building the tractor-bike, we build the seedling tray. But with the change in dimensions of the back wheels and sprocket, we now have to adapt the length of the connecting parts between bike and transplanter. To solve this problem, I feel lost without my computer, I cannot manage to visualise the best bay to rearrange the parts so that they fit together. Luckily, we move on to another problem, which gives me more time for thinking about the former one.

Une fois boulonnées en place, ça tourne mieux, et en plus on peut faire des virages !
Once the wheels bolted on the axle, it moves more smoothly, and we can even take turns!

The bike moves, but is often derailing. Once all the parts have been fixed, it is better, but to guarantee parallel parts when no measurement is precise is a big challenge. We fix small pipes to the frame to align the chain with the sprockets, so it doesn’t derail from the back any more, but from the front! At least now we can pedal a bit further.

Avec l'artisan/ouvrier qui a fabriqué la majeure partie du Ricycle :)
Me with the manufacturer who built most of the Ricycle!

Work in progress…

Maintenant, il ne reste plus qu'à fixer le mécanisme de transplantation... et ça va pas être une mince affaire !
Now, we only have to attach the transplanting mechanism! Well, only… it will definitely be a difficult task!

In four days, we managed to build a mostly functional bicycle-tractor! Now that the biggest part of the job is done, the most challenging part stays ahead: linking the transplanting mechanism and make it work!

Giving a hand

With the Cybathlon going on, showcasing the cutting-edge prosthetic technologies, I couldn’t help thinking about the low-cost available options when I met a kid with several limb differences in my project village in India. What if we could (litterally) give him a hand?


The e-NABLE community has been designing, printing and fitting hands and arms for children with upper limb differences. So far, they have solutions for hand- differences (missing fingers, functional wrist) and forearms (no functional wrist, functional elbow).  All of the prosthetics are open-source so that anyone can print their own and edit the design, allowing constant improvement and adaptability.

Isabella design:
Isabella design: “This device was created for those that have a functional elbow and a considerable amount of forearm but no wrist or not enough wrist/palm to power a wrist driven device.”

Here is a webpage listing all their released designs.


The UnLimbited Isabella and Alfie designs have fingers powered by elbow motion. If the bearer of the prosthetic bends their elbow, the fingers will close, enabling them to grasp objects.

The prosthetic is printed in PLA to allow thermoforming, and is fully scalable to be adapted to the recipient.

UnLimbited Alfie prosthetic fitted to a happy girl.
UnLimbited Alfie prosthetic fitted to a very happy girl, Sophia. Picture from Thingiverse.

Source files on Thingiverse

Look at those smiles when they try their new arms on for the first time!

3D-printed hands

Before learning about the 3D-printed arms, I knew about the Robohand initiative and 3D-printed hands. Here are some designs:

Cyborg Beast

Cyborg Beast 3D-printed hand from Thingiverse
Cyborg Beast 3D-printed hand from Thingiverse

Basic Robohand

Basic Robohand design from Thingiverse
Basic Robohand design from Thingiverse

DR2: Cycling in Mud

Those who know I’m crazy about bikes might also know that before starting my engineering studies, all I did was cycling around and rolling myself in mud in mountain biking competitions. I have some very fond memories of a French Cup race where there was so much mud people would slide down the slopes on their backsides. Now that is what I call mountain biking!

Sliding down hills of your bike is one thing, but how do people actually ride through mud? With the right technical equipment and some driving skills it is possible to race through mud fields. Lets analyse how the champions do it!

Riding through mud

See, those who go furthest usually have high rpm, and go as straight as possible. The bike has narrow tyres with high spikes to avoid building up mud. They shift their weight to the back of the bike to have more traction.

For those who fail, they will most likely get their front wheel stuck of slide on their sides. Some unfortunate riders will get their bikes stopped by mud blocking the wheels in v-brakes or frame.

If this kid can ride his bike through that clay puddle, than certainly we could get a bike tractor through a rice field, couldn’t we?

The physics behind it

Estimation of cycling power

Power is the rate at which energy is used (energy over time) and is measured in watts. In cycling, energy is expressed in terms of work (such as how hard you have to work to ascend a climb). It’s a constant snapshot of your work rate at any given moment. […] A better measure, especially on climbs, is watts produced per kilogram of body weight.
How much better are Tour riders than the rest of us? A contender for the overall classification can produce just above 6 w/Kg on major climbs of the race. By comparison, a domestic pro could manage a best of 5-5.5w/Kg; a good, competitive amateur or masters’ racer can put out around 4w/Kg and an untrained person would struggle to produce 2.5w/Kg.



Design Research: Pedal-Powered Tractors

When speaking about agriculture mechanisation, we immediately thing about huge tractors, pumping up oil and spreading loads of bad chemicals on the fields.
Some farmers have taken the problem the other way around, and instead of relying on heavy machinery for light operations, they started building their own machines, powered by humans, not petrol. I have listed the farm hackers I’ve found on this post.

If you drive a car, you’re dragging a lump of metal that probably weighs 10–20 times as much as you do wherever you go. What a waste of energy! Go by bike and the metal you have to move around with you is more like 6–9kg for a lightweight racing bike or 11–20kg for a mountain bike or tourer, which is a fraction of your own weight. Better efficiency means you can further on the same amount of fuel. According to the classic Bicycling Science book by David Gordon Wilson et al: “A racing bicyclist at 32km/h (20mph) could travel more than 574 kilometers per liter (1,350 miles per US gallon) if there were a liquid food with the energy content of gasoline.” Explain That Stuff

The Culticycle

Farmers from the Farmhack community have developed this pedal-powered tractor to reduce the usage of their tractors. It is built out of bike parts and standard metal pieces, and features a belly-mount for interchangeable cultivating implements (weeder, seeder, etc.)

The Culticycle, a pedal-powered tractor developed by Farmhack
The Culticycle, a pedal-powered tractor developed by Farmhack
Description: A pedal powered tractor for cultivation and seeding, built from readily available lawn tractor, ATV, and bicycle parts. Speed is 3 – 4 mph depending on choice of gearing and pedaling speed. Better for operator’s body, less soil compaction, no fuel use, cheaper than a tractor; easily adaptable to specific needs

Problem Statement/ Functional Need: Cultivation only requires the movement of small amounts of soil, therefore very little power output. Small tractors are hard on the body of the farmer, cause soil compaction, cost large amounts of money, are complicated to fix, and provide significantly more power than is needed for many seeding and cultivating jobs. Cultivation with a pedal powered machine provides sufficient power, a less physically damaging experience for the operator, and is more environmentally sustainable.

The Aggrozouk – previously known as Bicitractor

The Aggrozouk has been developed by the French-speaking collective of farmers l’Atelier Paysan, who develop tools for organic farming. The machine has been further developed during the POC21 eco innovation camp. They held a prototyping workshop in early 2016 so that the farmers could build their own Aggrozouk tractor while learning the skills to repair and upgrade it.

The Aggrozouk pedal-powered tractor with electrical assistance
The Aggrozouk pedal-powered tractor with electrical assistance

PR6: Open-source and robots for agriculture

Open-source Agricultural Projects and Hackerspaces

La notion de brevet, donc de notre point de vue de confiscation au profit d’individus ou de groupes, est contraire à notre volonté de contribuer à la production et la diffusion de Biens Communs. Nous estimons que la créativité est histoire de cheminements, d’influences, de rencontres, de glanages, bref, par essence d’une richesse collective, humaine, d’un génie créatif que nous avons choisi de ne pas garder pour nous. Toutes nos réalisations sont donc diffusées sous licence libre, pour une libre adaptation, pour que les machines et matériels soient vivants, appropriés et appropriables. L’Atelier Paysan

Aggrozouk/Bicitractor pedal powered tractor with electrical assistance
Aggrozouk/Bicitractor pedal-powered tractor with electrical assistance

L’Atelier Paysan: building open-source farm machinery for organic farmers, developers of the Bicitractor/Aggrozouk pedal-powered tractor with electrical assistance

TechAguru: building small-scale open-source solutions in the Philippines to bring precision agriculture to farmers.

Hacker Farm: Japanese countryside farm and hackerspace, have a project on rice paddy water level monitoring.

Farm Hack: Worldwide community of farmers that build their own agricultural machinery, famous for Culticycle pedal powered tractor

Good Tech: Community of makers testing prototypes in a realistic environment. Accelerator program for “bringing sustainable tech into mainstream“.

Fietswieders: “a lowtech agricultural machinery, based on open source to make the life of (organic) farmers more fun!” by Dutch community, using bicycles to make machines suited for work lying in prone position (facing the ground).

Obviously better technology will only make a difference if it´s actually used by many people. Therefore scaling new products is a key aspect of what we do. GoodTech

Culticycle pedal-powered tractor
Culticycle pedal-powered tractor

Small Farming Robots

For decades, farm machinery has targeted industrial-sized farmers, underpinning the “get big or get out” ag model of consolidation. Now, the miniaturization of farm machinery may be the ag-tech counter-trend that actually encourages smaller, more diverse farms.

Weeding robot Naio
Weeding robot Naio

Even in poorer nations, farm labor is not always available, as people are flocking to cities in increasing numbers. Which brings us to HelloTractor. Calling itself the Uber of Farm Machinery, this startup based in Washington, DC and Nairobi, Kenya, allows farmers to request farm machinery, just as you might “hail” a car with Uber. HelloTractor’s delivery system is tied to its own small, smart tractors, which monitor usage and location for the security of the owner. Owners can help offset the cost of their purchase by renting it out. And because labor shortages on farms can lead to poor harvests and lost income, the wider availability of these size-appropriate machines can help whole communities grow.

Yes, big machines may have allowed a single person to farm miles of land. But they also created farms low on diversity. Small machines could not only help large farms to become more diverse and ecologically sound, they can be a huge help to small, diversely planted farms that suffer from too little machine solutions to help them. Source

Rowbot: robots for small-scale agriculture in corn farming

Agribotix: Drones for agriculture

Naio: robots for small-scale agriculture

Prospero: agricultural hexapod robot prototype, designed to work in swarms

PR4: System of Rice Intensification

The System of Rice Intensification – known as SRI – is a set of agricultural methods for optimising resources developed in the 1980’s in Madagascar. SRI aims to increase the rice farming yields while reducing inputs such as water, seeds and fertilisers, so that rice farming becomes more profitable to the farmers. Unlike the techniques inherited from the Green Revolution, SRI does not depend on genetics engineering and chemicals. Depending on the countries and conditions, SRI can improve the yields from 20% to 100% compared to conventional farming.

SRI involves significantly reducing the number of rice seeds planted, transplanting them to the fields when they are much younger than usual, using different amounts of water at critical times of their growth cycle, and improving soil conditions with organic manure. […]

In Tamil Nadu, farmers are experiencing similar increases and are paying less. “Our chief minister’s aim is to get double the yield and triple the income of farmers using SRI. Traditional farmers use 30kg of seeds [compared with] 3kg by the SRI method” The Guardian

SRI Characteristics

Transplantation: In SRI, young seedlings are planted individually, instead of transplanting several mature seedlings. Thus, the seed requirements are 90% lower than in traditional rice farming. The seedlings are planted in a matrix pattern (25x25cm) with a wide spacing between the rows to allow for an increased exposure to the sun and wind, as well as better access to soil nutrients. The roots have to be intact and the seedling is planted shallowly (1-2cm instead of 3-4cm).

 With SRI, optimum spacing can be up to 50 x 50 cm (4 plants per sq.m) for very fertile soils. Best spacing is a function of soil fertility. Source: SRI Issue 6, 2009

Water management: In SRI, the fields are not permanently flooded, reducing the water consumption of the fields by up to 50% and decreasing methane emissions by oxygenating the soil. The cracks in the soil occurring when the fields dry allow for better oxygenation of the soil promoting root growth. When the fields are not permanently flooded, the plants also grow stronger stalks, which makes them more productive and resilient to bad weather conditions.

Weeding: SRI calls for regular weeding with rotary tools, which also aerates the soil.

Fertilisers: In SRI, the use of synthetic fertilisers and pesticides is reduced or banished while promoting the use of organic fertilisers such as green manure and compost. The nutrients are obtained by micro-organisms.

Sources SRI India / TNAU / various


SRI is not a standardised technological method, it is adapted from country to country and adapted to local conditions. It has lead to larger yields with lower inputs in many farms, although in other places it has been abandoned as not more efficient than traditional methods.

Adoption: As with any new technology and radically different approach to farming, farmers have to change their habitual way of farming for SRI, which makes it difficult to adopt.

Labour: In the beginning, SRI might require more labour for transplanting and weeding. Once the farmers are experienced in SRI patterns it does not require more labour, some farmers report requiring less labour.

Automation: The existing machines for rice transplanting are made for transplanting several seedlings at once, in a narrow matrix.
If a machine adapted to SRI can be built, the adoption would be easier as it will decrease labour requirements.

Success stories in Bihar (2013)

SRI history in Tamil Nadu

The Department of Agriculture of the Government has included SRI in all existing and new schemes funded by Governement of India that focus on increasing food production. […]
Further trials conducted at Aduthurai and Thanjavur showed that adopting all SRI components resulted in 48.8 per cent higher yield at Aduthurai and 35.8 per cent higher yield at Thanjavur when compared with conventional cultivation. A systematic study also showed that among the SRI components, the major effect was by weeder-use followed by single seedling per hill. SRI India


SRI has been introduced in Tamil Nadu in the early 2000s. However, a similar practise has been developed by local farmers over a century ago. Single-seedling methods and the Gaja methods have been known by farmers for a long time. The main difference between Gaja practises and SRI is that the seedlings are transplanted when mature in Gaja compared to young seedlings in SRI.

Today SRI is known to many rice farmers of Tamil Nadu as ‘Ottrai Natru Nadavu’ (single seedling planting). This recognition has come through SRI. But, to our surprise,
we find that single seedling planting was known100 years ago in Tamil Nadu. […] Single-seedling cultivation appears to have been developed by Mr. Aparanam Pillai (location not known) during 1905-06 season, and the  Gaja planting method,
which  also  included  single-seedling planting, was apparently developed by Mr. T.S. Narayanasamy Iyer of Thirukkaruhavur in 1911. Source: SRI Issue 6, 2009