Rezultaty
Three submersible prototypes of HAB Buoy have been built and successfully installed at three partner locations (HAB1 at Intecmar in Villagarcia, Spain; HAB2 at LBM in Trieste Italy and HAB3 at the Marine Institute Galway Ireland). A fourth development version at Plymouth University, shared with Plymouth Marine Laboratory, is being used as the test-bed for optimisation and performance assessment.
Staff at all sites have been trained in the operation and maintenance of HAB Buoy. Phytoplankton samples have been analysed using HAB Buoy as part of the commissioning process. The recognition software was finalised but is undergoing a process of continuous refinement and revision.
Partners have agreed to continue to operate HAB Buoy after the cessation of the CRAFT contract. As part of this agreement HAB Buoy will undergo sea trials in the spring of 2006, when the spring bloom provides a significant increase in plankton populations at all four sites. The consortium has received two sales enquiries, from the US and Peru.
The HAB system images everything in each seawater sample including detritus and non-harmful plankton. Each specimen is assessed and rejected if not relevant. The relevant specimens are then further analysed to decide on their species label. This label and the image of the specimen will be passed, via the Internet if connected, to
i) shellfishery staff,
ii) government health laboratories and
iii) water quality SMEs. The instrument will enable shellfishery staff to have advance warning that HAB species are present and their abundance. Government laboratories will be able to use the information to focus their resources more effectively. HAB Buoy will create a new market for water quality monitoring SMEs, who will be able to deploy and operate HAB Buoy in aquaculture industries in Europe and through out the world.
HAB Buoy is deployed and in routine use. It is still being trained although recognition of smaller species is expected to be difficult. However, the machine was designed for high throughput (375ml per hour) and probably can be targeted at monitoring changes in microplankton populations and hint at the distributions of small microplankton (20 <> 50 micron). It probably also has an important use as a monitor for shellfish larvae, to allow direct monitoring of mussel spat for example, as well as distributions of feed stock for mussels - to support predictions of shellfish growth rates. A wider role for HAB Buoy is possible with its capacity to image and recognise mesozooplankton. To this end, Williams and Culverhouse will be presenting HAB Buoy at a Globec workshop on automated recognition of zooplankton in November 2005.
HAB Buoy was designed to be modular and so variations with higher resolution optics (for smaller specimen sizes) and larger flow cells (for mesozooplankton) can easily be added in the future.
A major milestone has been achieved at the close of the HAB buoy project. A microplankton monitoring instrument has been designed, built and deployed for routine use at Marine Laboratories. The first steps have been taken towards complete automation of labelling plankton specimens through HAB Buoy. Opportunities for commercial exploitation exist already for laboratory use, but once tested in situ HAB Buoy will be a state of art next generation monitoring instrument that will lead the way for wide spread adoption of biological monitoring instrumentation. Tools such as this are essential for marine science and also for the successful implementation of the Water Framework Directive.
The data used to train HAB Buoy relies on the expert knowledge of skilled marine scientists.They establish the species label for each specimen, and hence define the gold standard training set.
This data is expensive to obtain, but is essential to the successful training of a learning machine like HAB Buoy.
Thus, this know-how constitutes a result.
Wyszukiwanie danych OpenAIRE...
Podczas wyszukiwania danych OpenAIRE wystąpił błąd
Brak wyników