Everything you wanted to know about Flowertokens but were afraid to ask — Part 2: Physical infrastructure

4 min readJul 20, 2018

Flowertokens is a prototype: a small test case with which we hope to provoke a broader discussion around the tokenization and verification of natural commodities. Even if it works in its own right as an experimental micro-economy in both technical and economic terms, the project functions mainly as a visualisation of certain mechanisms. It demonstrates what is possible (and more importantly, what is not). This approach is very much dependent on our different narratives of the future, and the scenarios related to them.

In the early stages of the project, we discussed at length which plants would be most suitable for our purpose. In our narrative, the plants are substitutes for any living organism (such as a tree) or even an entire ecosystem. Yet they must also act as objects of identification: as distinct collectibles.

In the end, we decided on flowers (and against crops such as lettuce or tomatoes) because it seemed easy to attribute to them an abstract value which differed from their functional value. A secondary (more practical) reason was because of the speed at which they grow — and thus the speed at which the different metrics of their corresponding tokens would also change.


After that was set, we needed to find a species of flower that (a) would bloom in a variety different colours, in order to have easily distinguishable plants, (b) would not exceed a certain height, so that 100 flowers fit on the shelf, (c) could still be planted in the current season, and (d) was able to be grown inside. To avoid the mistaken analogy between the tulip mania and the current cryptohype, and because they are not well suited to be planted indoors in spring, we decided against using tulips. Bulbous plants (such as tulips) do however grow vigorously and flower within a short time; thus we initially experimented with growing small gladiolas and freesias. Unfortunately, the gladiolas grew to be too tall, and the freesias didn’t grow at all because of the temperature. Time was also a factor in our ‘test runs’, as testing the phases of a given plant’s growth would have taken an entire season (one year).

We then consulted a botanist, who advised us to switch to annual flowers grown from seeds, as this would give us the best chance of flowers blooming this season. However, there were certain downsides associated with these plants: their slow growth and the associated problems with the Oracle (see up coming — Part 3 for more).

The flowers we chose are small annual dahlias (Dahlia x Hortensis). The seeds were pre-grown in special soil before being repotted and placed into the grow-rack as seedlings on July 8th.


The physical installation is built for — and set up within — Trust, a co-working and event space in Berlin. The system had to accommodate 100 plants as well as adapt to the spatial conditions. The requirements were to develop a cost-effective system that would be able to provide the best possible care for the plants and run at least semi automated. Since the entire information feedback system would rely on a visual input (camera monitoring), we rebuilt the space inside Trust as a 3D model, and simulated different options, shapes, and camera setups.

To cut costs, the actual system is made of modified steel shelving systems, coated plywood, and drywall construction elements. Even though it is situated next to the window front, the sunlight is supplemented by 20 full-spectrum lights with a total output of 180 watts controlled by a time switch.

During the planning and construction process we quickly realized that not only would everything become very expensive, but that watering the plants by hand would also take a lot of time. We therefore decided to install a custom microdrip system, enabling us to partially automate the process of watering the plants. Surplus water is drained off and filtered before being fed back into the watering cycle. At the time of publication however, since the plants are currently very small, they must be watered by hand; the microdrip system has yet to prove its worth.