Skip to main content

A novel smart trap station as an Internet of Things surveillance solution to remotely count and identify the species of disease-carrying mosquitoes

Periodic Reporting for period 2 - REMOSIS (A novel smart trap station as an Internet of Things surveillance solution to remotely count and identify the species of disease-carrying mosquitoes)

Reporting period: 2017-02-01 to 2018-01-31

With global travel and the efficient flow of goods and commodities, the EU also experiences an unintended influx of animals and plants that are not native to Europe. Some may become established, a process which is often facilitated by climate change. Such invasive species may not only pose ecological problems by out-competing native species, but can also pose a significant public health threat, because they are potential vectors of exotic diseases. The public health system with its surveillance networks and its risk assessment needs to adapt to the evolving situation. One aspect is the surveillance of potential vector mosquitoes. Even using the most efficient trap, like the BG-Sentinel from BIOGENTS, this is time consuming and work intensive. Data are typically available only after days, often weeks. The surveillance efforts still represent more than 95% of mosquito surveillance costs. REMOSIS addresses this problem. It makes the surveillance traps smart, automates the detection and counting of mosquitoes and provides reliable information electronically almost in real time, and it enables highly focused, efficient vector and disease control activities, reducing operation costs by 80%. This is a disruptive leap in mosquito surveillance.

The project successfully reached its objectives, specifically:
1- The development of an advanced trap station with mosquito counting module and remote management tools. This system (BG-COUNTER) is already commercially available, with more than 200 currently sold.,
2- The developments of a further advanced system to also distinguish mosquito species of interest. Out of three sensor designs that were developed and evaluated, one, the BG-EYE is now moving on to production to meet customer orders for summer 2018.
3- An advanced WSN-Wireless Sensor Network (nodes and gateway) was prototyped.
4- With a cloud application specific for Mosquito surveillance, the devices can be managed, the data they collects can be downloaded and displayed with a maximum delay of 15 minutes.
The work in the REMOSIS project was divided into different 5 Work Packages.

WP 1: The project was coordinated, monitored and managed in regular emails and conference calls. Every 6 months a personal or remote meeting was organized with representatives of the consortium partners. Results and findings of the R&D process were continuously screened for relevant IP, and the commercial and scientific potential was evaluated for its economic potential, patentability or publishable scientific interest. Knowledge needs were identified and ways to gain access to needed know-how were established, enabling the industry partners to effectively exploit the results achieved in the project.

WP2: A web page of the project was created and maintained. The project and product developments were disseminated in:
- 13 scientific congresses and 25 business events
- pilots with 9 mosquito control entities
- 2 peer reviewed articles and 7 posters/oral presentations
- demos in 13 workshops
These activities led to securing accumulated pre-orders and orders of more than 200 and rising. Currently partners BIOGENTS and IRIDEON are working together to satisfy the requests of customers willing to pay for the final results of the project, with first deliveries planned for June 2018: time-to-market of REMOSIS <6 months.

WP3: TEIC produced the TRL7 optoelectronic sensor prototype for species classification, which was extensively and successfully tested in laboratory with live insects. The sensor is capable of accurately differentiating mosquitoes from other insects, as well as males and females. Accuracy levels were >90% for mosquitoes actively flying though the sensing area. After the integration into a suction trap and simulating field conditions, the classification fell to 50-55%. We identified several opportunities to raise the accuracy further by incorporating different sensor types, culminating in three improved species ID prototypes. In the second year of project, two of these versions were selected to proceed on to the trials and pilots to achieve TRL8, and industrialization in WP5: the BG-EYE (design by BIOGENTS) and the DUET (design by IRIDEON). Several units of these were manufactured to serve the worldwide pilots.

WP4: The electronics of the final REMOSIS unit designed by IRIDEON have achieved TRL9. The electronics are optimised for mass production at competitive prices, with total control on the quality, compliant with international standards and CE mark. The server application, developed by BIOGENTS, consists of two domains: one with information on the REMOSIS device, and one with a dashboard, from which traps can be controlled and monitored, and data downloaded. IRIDEON developed an interoperable, standards-based middleware called SENSCLOUD, to be incorporated into the REMOSIS cloud application to make it IoT enabled. .

WP5: The BG-COUNTER reached TRL9 during the second project half. Demonstrations were performed in several countries in Europe, America, Asia and Oceania. Field trials were conducted at several sites in Germany, USA and Australia. The market acceptance was overwhelming, leading to the placement of more orders and an increasing interest.
The species ID prototypes, BG-EYE and the DUET, were trialled and piloted with mosquito control experts and authorities in the Iberian Peninsula, Germany, USA and Asia.
The DUET managed to reach TRL7 and was close to reach TRL8, with an accuracy of >90% for free flying mosquitoes and around 65% when combined with a suction trap. Due to adverse weather conditions in the Iberian Peninsula, not enough field data could be collected to reach a consolidated TRL8.
The BG-EYE was demonstrated and piloted outside of Europe, namely in the USA and Asia. Enough field data was captured to enable this prototype reaching TRL8 with a mosquito classification accuracy above 70% when used with a suction trap. These values further increased with improving the AI and machine learning algorisms used. With the end of the project the industrialization process commenced, making this unit reach TRL9 in spring 2018, and meet orders placed to be delivered to the early adopter customers in June 2018.
The current surveillance of nuisance and vector mosquitoes relies on the manual collection of the catch from mosquito traps and a manual counting and identification of the catch. This is time-consuming, expensive and yields data only with a considerable time lag (days to weeks). The REMOSIS devices automate the detection and counting of mosquitoes and provide reliable information electronically from as many locations as needed simultaneously and almost in real time (every 15 minutes). It thus enables highly focused, efficient vector and disease control activities and an instant quality control of the taken measures. The collection of environmental data further enables the users to better understand and predict environmental influences from the acquired data, and to better plan control activities. The impacts of the project include:
- Predicted sales of €5,2 million over 5 years, with NPV of €3,2 million and ROI 1,19;
- Elimination of costs for manual inspection of traps with a solution 3 times cheaper than reference SoA;
- Improving surveillance and control of disease-carrying mosquitoes and thus reducing the burden mosquito-borne diseases, as well as making mosquito control activities more efficient and environmentally friendly.