I am currently working on the topic of sustainable, community-based food systems. Such food systems are based on agroecological principles. They are also collaborative and decentralised. As a computer scientist, I am interesting in the role that technology can play in such systems. I am aware, however, that technology can only solve part of the equation. That is why he also has a strong interest in the "Commons" as a concept to deal with many of the socio-economic questions related to food systems.
Peter Hanappe studied electronic engineering at the University of Ghent, Belgium. He wrote his Ph.D. thesis on real-time music and sound environments at Ircam (Centre George Pompidou, Paris). When a researcher at Sony Computer Science Laboratory in Paris he has worked on new modes of content creation and distribution that involve the participation of (online) communites. He now focuses on projects in the domain of sustainability. He was involved in setting up a collective system for monitoring noise in urban environments (NoiseTube) and in building critical components for large-scale climate simulations in collaboration with the U.K. Met Office. Most recently he founded the P2P Food Lab to experiment new technologies for agroecology and community-based food systems.
P2P Food Lab aims to develop new socio-economic models for sustainable food production. I am seeking solutions that promote the local production, distribution, and composting of food. The hypothesis is that with the use of today’s information technology, our understanding of the Commons and the insights gained from the Free Software movements it must be possible to design food production chains that are more ecological, decentralised and collaborative yet as productive as today’s food industry.
P2P Food Lab looks at the combination of agroecology and technology, and at the socio-economic issues that arise when those two come together. To explore the boundary those two, I have started several concrete projects:
We work on an automatic weeding robot for agroecological farms and bio-intensive cropping techniques. One of the characteristics of bio-intensive croppping techniques is growing many varieties of vegetables on small surfaces. The plants are densely spaced, allowing high productivity on small surfaces. Due to this geometry, the work in the fields is mostly manual. (See John Jeavons, but also Eliot Coleman and, more recently, Jean-Martin Fortier.}. In these conditions, weeding is done mostly manually. It is a tough and fairly uninteresting job.
The robot uses cameras to detect the plants in the field and uses a computer numeric control (CNC) to move a weeding tool accurately between them. The device could be extended to do other tasks, like precision seeding (like FarmBot), transplanting, and - possibly - harvesting. The robot has a set of sensors that could be used for other purposes, such as early pest detection or phenotyical analysis of plants. This data can be stored in online servers to build maps of the fields, analyse crop health and harvests, and other services.
The project is still in an early stage. We will post more info on the Web site.
Entry page: https://p2pfoodlab.net/WeedingRobot.html
CitizenSeeds is a participative project for amateur gardeners. They are asked to share their experiences of growing a given set of seeds. We prepare and send out the collection of vegetable seeds to the participents. They can plant these seeds on a small surface (maximum 1.5m2) and then take a picture of their plants once a week. The the online project page collects all the photos. One of the goals of the projects is to use gardening and online tools to raise awareness about agroecological food production techniques and to build up a social network around local food production.
The CitizenSeeds build upon the experience of the Starter Kit.
The capsules with the seeds as they are sent out.
A screenshot of the web site. The photos are displayed in a matrix. A row corre- sponds to the contributions of a participant. A column correspond to one week. Participants can upload pictures of their vegetable bed, but also of individual plants.
The participants can use a FlowerPower to register environmental data. The main project page includes a summary of the environmental data. The full details of the environmental data is shown in a pop-up window.
The Starter Kit is an Internet-connected greenhouse to help people grow vegetables in urban and peri-urban areas. The platform consists of a small greenhouse (1.2m x 1.2m), a sensor box with wireless communication, an online community web site, and a collecion of seeds.
The first version of the greenhouse.
The sensor box inside the greenhouse measures the temperature and humidity (inside and outside), and the sunlight. A webcam takes pictures of the plants growing inside the greenhouse. The designs of greenhouse and the sensor box are open and can be easily built by DIY’ers. We aimed to make the equipment as low-cost and as adaptable as possible in order to reach a wide audience. For the sensor box, we use an Arduino for the measurements and a Raspberry Pi for the wireless communication and webcam.
On the community web site, participants can interact with each other to ask questions, provide tips, or simply annotate their activities in their online notebook. The gardeners can visualise each other’s sensor data and notebooks. Submitting new notes is simple. Hashtags can be used to categorise notes and images can be attached to each post. Several hashtags have a predefined semantics and transmit the posts to specific message channels such as, for example, the ‘questions & answers’ or ‘tips’ channel. Our aim is to use this common platform -greenhouse, sensors, web site – as a tool for citizen science projects in the field of vegetable gardening.
One of the objectives was to develop a citizen science platform: gardening enthusiasts can use the Starter Kit to coordinate scientific experiments and study the effects of plant varieties, soil or meteorological conditions on the productivity. For example, by using the same seeds in different locations, or by associating different species together, we can obtain data about the efficiencies of different strategies. Quantitative data can be obtained by analysing the images, the sensor data, and the notes submitted by the participants.
In 2007, I initiated a collaboration with the Physics Department of Oxford University and the U.K. Met Office. The Department of Atmospheric Physics at Oxford was launching a new major scientific experiment: the Millennium Experiment. The goal of the experiment was to reconstruct the climate of the past 1200 years using the U.K. Met Office's FAMOUS/HadCM3 climate model.
An interesting aspect of the Millennium Experiment is that the University of Oxford relies on a network of volunteers (ClimatePrediction.net) to perform the computation on their home computers.
The computer modelling is part of a larger European research project European Climate of the Last Millennium that groups 38 research institutions. The majority of the groups work on the reconstruction of the past climate using historical documents and natural archives.
© The Millennium project.
The Sony Computer Science Laboratory Paris is not a partner in the European Millennium project but collaborated closely with the University of Oxford and the U.K. Met Office to evaluate the performance and the energy consumption of the FAMOUS climate model.
This collaboration has led to:
We ported the FAMOUS climate model to the PlayStation 3 in the hope that owners of the Sony's game console could participate in the Millennium Experiment. The video below shows the model running on the PS3.
A short presentation of the Millennium experiment by Hiro Yamazaki.
The Millennium experiment on the ClimatePrediction.net web site.
The web site of the European Climate of the Last Millennium.
Installation and start-up on the PlayStation 3:
The new grahical interface:
The goal of this project is to develop the most energy-efficient and low-cost solution for performing large-scale scientific computions. My work is is currently focused on volunteer computing.
A volunteer computing network consists of a set of home computers that are connected to the Internet and that are configured to perform scientific computations when they are idle. Volunteer computing is well-adapted to handle large "pleasantly parallel" problems.
Because many computers are potentially involved, the total computating power can be enormous. But, for the same reason, the total energy consumption of the network can be enormous, too.
We have found that the energy needs of volunteer computing networks can be reduced significantly when the power saving feature of recent processors is used intelligently. However, to achieve this reduction, the operating system must provide additional support. Unfortunately, this support is not available.
I am therefore studying how to circumvent this problem and how to keep the machine artificially in a low-performance state. Using this technique, it is still possible to reduce the energy footprint of volunteer computing.
My current goal is to integrate new tools to measure the energy consumption and the CPU's performance-state in the BOINC framework. This may not only help in reducing the energy, but may also help extend volunteer computing to mobile devices.
The starting point of the Phenotypes / Limited Forms project is the archive of photographer Armin Linke, which contains over 6000 images.
We were seeking a new form to present such a large archive in an exhibition space. We wanted to go beyond a static display and, instead, give people the opportunity to build their own interpretation and then leave a trace of this interpretation.
What we did is produce about a 1000 of Armin's photos on high-quality photographic prints and put the prints on three long shelves, 12 meters long. Visitors of the exhibition can just flip through the photos and simply enjoy Armin's rich archive. It's a bit like flipping through stacks of vinyl records in a record store. As people move images around, the wall itself becomes an evolving display of the archive.
Photo © Armin Linke
People can pick out the prints they like and carry a selection of photos with them. This selection can then be used to print a small book. We designed an editing table for this purpose. It is positioned in the middle of the exhibition space. Visitors lay out 8 of their images on the editing table. The table uses RFID technology to detect the images that are laid out on top.
Photo by ARTExplorer
When the visitors are happy with their selection they print out a paper copy using a thermal ticket printer. Before printing, people must enter a title for their book using a virtual keyboard on a touch screen in the table.
Photo by ARTExplorer
As soon as someone finishes a book, we're projecting the title of the book on the a wall in the exhibition space. We also show the titles of the books that were made by other people but that are related to the new book. The idea is to make people aware that creating a book is not an isolated event but adds a new point of view to the archive.
Besides the physical installation, we created a web site that functions very similarly but is accessible on-line. You can navigate through Armin Linke's archive and when you find photos that you like, you take them along with you by dragging-and-dropping them in a clipboard. As soon as you have a selection of images, you can start creating a book using the book editor. Simply drag-and-drop the photos from the clipboard onto the virtual pages of the book.
For 100 euros, it is possible to obtain a paper copy of it that is printed by Graphistudio in Venice using the same high-quality printing technique that is used for the prints in the installation.
More than 40,000 people have used the installation to create a book. All the books that have been created in the installation are available on-line.