For thousands of years bees have played a vital role in sustaining our ecosystem through plant pollination, and the humble bee has given us the sweet delight of honey. Professor Nikola Kezic, a specialist in bees at the University of Zagreb in Croatia, is creating a buzz by showing off another important function of bees; their ability to detect mines.
He has been a professor in the Faculty of Agriculture since 1985 and is close to retirement this year. His swansong is a four-year project working with the European Union's Toolbox Implementation of Removal of Anti-Personnel Mines, Submunitions and Uxo (TIRAMISU) and also Mine Action in Croatia (CROMAC).
It is estimated around 1.5 million mines were deployed during Croatia's battle for independence in the 1990s and landmines still cover around 667km² of the country. With Croatia's entrance into the European Union on 1 July, de-mining is at the top of the political agenda. Professor Kezic tells Army-technology more about his research.
Grant Turnbull: Why do bees make good mine detectors?
Nikola Kezic: Bees have a well-developed sense of smell so they can distinguish different flowers; every flower has a particular smell. The bees are often going from one flower to another to collect pollen and nectar for food.
Their sense of smell means they can recognise a particular flower and carry on collecting food from it over a period of time. We thought this ability could be very useful and began researching their ability to detect the smell of TNT.
In 2002, we started experiments looking at the bee's proboscis reflex (when the bee extends its 'tongue' for a sugar solution) and training the bees as a colony. It was a challenge at first to get the bee colony to work as we would like, individual bees are much easier.
GT: How easy is it now to train the bees?
NK: It's very simple. We place a glass containing a sugar solution in soil mixed with traces of TNT. The bees are attracted to the sugar solution and need to pass the TNT to get to it.
We are not mixing TNT with the sugar solution. Once the bees learn they can find food in an area where there is the smell of TNT, they follow the TNT everywhere. We just need an amount of TNT that would normally be found on the surface of a mine.
The training period is very short and they learn fast, in just three or four days you have well trained bees. To be effective I have to isolate the colony, I train the bees in mesh tents where only I can feed them and it is only food connected to the TNT smell.
Once they are taken out of the mesh tent they then try to find the TNT smell because they associate it with food.
One of the big improvements at the moment is that we are able to keep their interest in TNT over several days.
GT: how has your work been received by the international community?
NK: Mines are a big problem in Croatia, thousands of our people have died and suffered terrible injuries because of them. I want to help in any way I can. In the beginning I was just experimenting, but now I have a project which is supported by the European Union.
I have funding from the EU's TIRAMISU project which means I need to get results. My work with TIRAMISU began last year and we are now in the second year of a four year project.
GT: At what stage is your TIRAMISU research?
NK: I am nearly at the end of my experiments now and need to write all the reports for the project. But this is the first stage; the next stage is the validation phase before they start to implement the research. I need to write an implementation procedure and they need to verify the results, which is done by an independent commission and it usually takes one or two years.
I will carry on these experiments even when my work with TIRAMISU is done and I retire. Bees are part of my life, they are my profession and I like what I am doing.
GT: What challenges are you trying to overcome?
NK: The biggest problem is how to follow the bees when you release them into a test minefield. You have thousands of bees flying around and you need to be able to detect them while they are flying. Another group from the University of Zagreb are working on ways to detect the bees using airborne monitoring equipment.
The main equipment they use to detect the bees is an infra-red camera, when the bees are flying they produce heat and we can detect these hot spots using the camera. It's very difficult to distinguish them in the real environment, but with infrared you can see where they are landing.
GT: How do you see mine-detecting bees being used in the future?
NK: This technique, searching for TNT material and not for the mines themselves, can be used for quality control. You need verify if a particular field has been de-mined or not. It needs to be cleared of all mines but scientifically this is not possible.
When an area has been fully de-mined someone has to go over the ground again, checking areas at random just to prove it is de-mined.
This takes time, bees can cover all this area in one go and highlight areas that may need a second de-mining.
This method increases safety and quality control but we are not replacing any technique at the moment, we are just adding a new technique and a new possibility to prove it is really demined.
GT: Is there are a risk of being stung while de-mining?
NK: They don't attack people in the field, often if they sting they are just trying to protect their homes. People are not exposed to a high risk of being stung when they are near the minefield. Once you begin working with the colony, opening the colony, managing the bees then they can sting you from time to time, but it's normal and I must say healthy! Beekeepers have a good quality of life.
Freed from their tethers and from minute-to-minute human guidance too, today's robots are a far cry from their forebears. Dr Gareth Evans reports on how companies like Robotics Engineering Excellence (RE2) are working to equip military robots with greater levels of on-board decision-making capabilities - and perhaps, ultimately, full autonomy.
The latest round of grants under the US Department of Defense University Research Instrumentation Program lists Brain Computer Interfacing as an area of funded research.