For many years now, I have been creating experimental set-ups using sustainable beehives that have been augmented with camera’s, microphones, sensors and sensory processing algorithms to analyse the state of the colony, the quality of pollen and the behavior of the bees. These “Intelligent Beehives” are progressively linked in a European- wide network and the data is being made available online. Continue reading
We have been developing a monitoring device that is based upon the continuous monitoring of the colony’s buzz: a non-intrusive scanning device for controlling the colony’s health & development. We also have been adding video monitoring (outside and inside), which gives us a full spectrum of possibilities for colony monitoring and environmental surveillance.
As bio indicators, honeybees provide us with a constant stream of information on the environment (urban, countryside) on which they forage (activity, pollen, nectar). Diseases like colony collapse disorder and environmental problems like the use of pesticides could be analysed in a different way by monitoring and analysing the daily activity (audio, video) of several bee colonies over multiple years.
In our test station in Brussels city center, we have 2 beehives equipped with off the shelf-technology for monitoring bee activity at the landing platform (2 x video, outside and inside) and for monitoring the health and development of the colony by sound recordings of its activity (8 x audio, inside). The test station also hosts 4 non tech. equipped beehives which are usefull to make observations at the flighthole/landing platform and to compare these findings with the results of the digitally monitored hives.
Vincent was preparing the audio for the sound beehive: 4 electrets microphones and 4 piezo microphones with preamps mounted in the rooftop of the Warré beehive. We’ve put the charger for the preamps a couple of meters away from the hive, to avoid all EMF and to be as less intrusive as possible.
Our initial intention is to install the Asus computer (with debian) and a Mackie mixing panel. Later we decide to swap that setup for a more performative one: an 8 channel Prosonus soundcard, the Asus with Debian for recording and sending the files over the network to a NAS (network attached storage) hard disk.
We will record 4 times 3 minutes an hour, every :00, :15, :30 and :45. The 8-channel files will be archived via the computer & network on the NAS, the computer then compiles the .wav files into a stero mp3 and a selection of the most recent files will be streamed to the OKNO server for broadcast.
Vincent is installing the Prosunus soundcard and the Asus/Linus computer. He wrote a script to record every 15 minutes 3 minutes of sound on 8 channels. This makes 12 minutes on 8 channels per hour. The recorded files are send to the NAS (storage HD) in my studio. The last 8 files are compiled into a stereo mp3 and uploaded on the Okno server as a playlist. We are now testing the system during a few days, before the bees arrive.
The bash script makes it possible to manage the recordings from a distance, online. Which canals, how many times, etc… everything is modular. But the goal is to automate the system once we are sure about the perfect setup.
The bash scrip at server-side controls the recordings on regular intervals, the bash-script on client side synchronises the playlist of the last 30 minutes of recordings. There is also an online archive that can be consulted.
The computer needs to be powerful enough to record 8 channels simultenuously, and as well compile into mp3 format and stream the playlist.
Cables, connectors, piezo’s and electrets: all the connectors to the preamps are located in the roof, above the upper box. Thje multicable (8 microphones!) is 10 meters long and comes out of the opening at the side of the rooftop. The length of the cable is of no importance thanks to the preamps in the rooftop. Everything is water resistant.
All technology in the hive is on DC, so there is no EMF danger for the bees.
The electrets microphones are also located in the rooftop, as such the bees won’t cover them with propolis. The cinch connectors are located on the topbars of the highest box (no other possibility) – hopefully the bees will not damage them with propolis.
I decide to do another bee-sound-experiment. The fist one I did was in 2012 with the Transparent Beehive. Then, the focus was on exhibiting in realtime the sound of the colony. During talks and presentations I was making observations and linking them to the amplified sounds made by the bees.
This time I want to do it differently. I will record at regular intervals the hum of the colony and analyse it thoroughly afterwards. I also want to link the sounds with the environmental sensor data (temp, humidity, solar radiation) in the surroundings of the apiary, with the sensor data inside the beehive (temperature, humidity and vibration of the comb) as well with video images in- and outside the beehive.
For this setup, I will collaborate with Vincent Malstaf (sound engineer), Balthazar de Tonnac (computer scientist), Bob Motté (electronica engineer) and Bart de Boer (computer scientist Artificial Intelligence, complex systems, bio-acoustics).
The Raspberry technology offers us a not too expensive solution to build out a suitable lab-setup for this research. The data will be available on the opensensordata website. Continue reading
Ethology is the scientific and objective study of animal behaviour, and is a sub-topic of zoology. The focus of ethology is on animal behaviour under natural conditions, as opposed to behaviourism, which focuses on behavioural response studies in a laboratory setting.
Many naturalists have studied aspects of animal behaviour throughout history. The modern discipline of ethology is generally considered to have begun during the 1930s with the work of Dutch biologist Nikolaas Tinbergen and by Austrian biologists Konrad Lorenz and Karl von Frisch, joint winners of the 1973 Nobel Prize in Physiology or Medicine. Ethology is a combination of laboratory and field science, with a strong relation to some other disciplines such as neuroanatomy, ecology, and evolution.
With the Sound Beehive experiment, we have been building a laboratory to study the development of the colony through its own sounds. The buzz of a colony and its behaviour and conditions are quite related. It is possible to know if a hive is queenless or if an important amount of nectar has been collected simply by listening to it.
For this experiment we follow a systemic approach to raise understanding of the characteristics of the colony through relationships with its environment, through patterns discovered in the collected audio, video and sensor data, and through contextual observations. We study the bees as a re-generating network of actors (autopoiesis), all of them contributing to the organisation and well functioning of the colony, the super organism.
Specific hardware and software is developed in order to continually monitor the sounds on different spots in the beehive.
We upload our annotated video and audio data for public viewing in our opensource videodatabase pandora. All corresponding sensor data are publicly available on opensensordata.net. The information archive grows as more audiovisual observations and more sensordata are added over time.
Masatoshi sends me an USB microscope to start developing our pollen database. I will photograph pollen brought back by the honeybees, and also pollen found in the garden.
Masa will do the same, and over a while we hope to establish a body of materials, starting to do some machine learning and later do pattern recognition. Another possibility is to compare the microscope pictures with existing pollen databases. Continue reading