������ϲʹ�������

In a thought-provoking seminar Chris Palego explained his work on using radar technologies, along with Artificial Intelligence techniques, to understand and explore bee motion. His enlightening talk took place as part of the Engage series of lectures. The lectures give academics, researchers, and companies a platform to discuss research in computing, engineering, and design. His presentation drew together a lively, and diverse audience of over 20 students and staff members eager to delve into the captivating world of honeybees.  

Chris explained “The significance of honeybees as pollinators cannot be overstated. They are pivotal to the soft fruit industry, food diversity, and ecosystem stability. Because of their small size, and rapid motion, it is challenging to capture their motion”. Previous work had investigated physical sensors on the honeybees. These could be considered as being backpacks for the bees. But because bees are small, if they are to carry any sensors, any backpack sensor technology needs likewise be extremely small, to not affect their movement. He went on to say: “to this end, we have been investigating how we can use minimally invasive techniques and automatic algorithms to comprehend honeybee movement more comprehensively.” Their work extends beyond honeybees and holds relevance for other insects, animals, and even assets, but the profound impact of honeybees and bumblebees on agriculture, food production, and ecological balance underscores the importance of their research. 

Chris, in his presentation, shared his groundbreaking research techniques, which utilizes radar. Their novel radar approach eliminates the need for invasive physical tags on bees. In addition, the new radar approach circumvents any inconvenience of putting a sensor on a bee, but also offers machine-learning-driven readout of Doppler radar signals. Chris explained, "Doppler signals are relatively narrowband and do not require extensive processing power, unlike video footage analysis of bee movement. After initial video-based training, the system can autonomously count and classify bee behaviour, potentially triggering actions by beekeepers or farmers."��

The ongoing study explores the relationship between machine-learning readout and traditional radar signal processing in analysing bee behaviour. This research endeavours to answer a fundamental question: which approach is more effective? Notably, their technology relies on off-the-shelf and cost-effective 5.8 GHz modules, each costing less than £100. This cost-efficiency makes the technology suitable for deployment in polytunnel farms, particularly for tomato pollination using bumblebees. By leveraging this technology to monitor different areas or single-plant visitation, farmers can optimize pollination efficiency, potentially boosting crop yield and sustainability. 

The implications of this research are profound for the agricultural industry. The decline in pollinators, particularly bees, is a mounting concern. The use of machine-learning readout offers the potential for farmers to gain deeper insights into bee behaviour, thereby allowing them to optimize pollination techniques. This could result in more efficient and sustainable farming practices, playing a pivotal role in safeguarding our future food supply. The innovative research serves as a testament to the significance of technology in addressing pressing ecological challenges. 

Bethany Johnson, the secretary for the ������ϲʹ������� IEEE (Institute of Electrical and Electronic Engineers) Student branch, said

“It was an extremely interesting talk. Chris not only enlightened us about bee movement but made the technical content understandable”.