Periodic Reporting for period 1 - ePatriot (Evolved Sky Patriot – Phase 1 Feasibility Study)
Reporting period: 2017-07-01 to 2017-12-31
Until recently, UAVs were not considered a security risk and their characteristics (small, slow, low altitude) make them virtually undetectable by traditional surveillance systems. As commercial UAVs proliferate and the interest of criminal and terrorist groups, as well as ill-intended individuals, grows for the use of sophisticated UAVs to support illicit activities, new technologies and solutions are required to counter them.
RNCA responds to this market opportunity and growing societal concern by providing an affordable and high-performing automated and transportable solution, capable of effectively and efficiently detect, classify and track UAVs, to assist LEAs and OCIs in maintaining public safety and security.
The Evolved Sky Patriot (ePatriot) uses optical and thermal video (allowing day and night surveillance), applying advanced video algorithms fully developed in-house and software functions with high levels of automation. Based on passive sensors (cameras), ePatriot can be used even in areas that highly constrain radio-frequency emissions or other active means, thus presenting broad and global business opportunities that address a major customer need.
The ePatriot1 Feasibility Study, funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 774256, allowed RNC Avionics to conduct a thorough assessment of the scientific and technological (S/T) viability of the innovation, as well as the economic and financial feasibility.
- Market Analysis
- Application Scenarios
- Threat Analysis
- Counter-UAV Technology Screening
- Potential Partnership Agreements
- IP management and patenting
- Standards, Certification and Regulation process
- Marketing and Communication
- ePatriot Cost-Benefit Analysis
- ePatriot Sales Forecast and Route-to-Market
- Organisational and Management requirements
- ePatriot Product Roadmap
As part of this feasibility study, RNCA was able to conduct rapid prototype development to test (in laboratory or simulated environment) the above obtaining the following results:
- RNCA adapted and tested video algorithms using video collected from ultra-high definition cameras (i.e. Basler acA4112-20u) delivering 23 frames per second (12.3M pixels). The algorithms were tested in a high performing desktop computer. After proper adaptation and optimisation, the video algorithms were capable to perform detection and tracking in real-time.
- RNCA conducted a literature review for the incorporation of dynamic learning techniques (that is, learn new previously unknown object types while online) in ePatriot. The most promising techniques consisted in incremental concept learning using hierarchical classification and deep convolutional neural network demonstrating that incremental training is possible.
- RNCA developed a ""clover"" system that encloses 3 cameras. Each camera has a 30º degree coverage, thus the clover system in total covers 90º degree coverage. By connecting 4 clovers, RNCA can deliver the full 360º degree coverage over an area.
RNCA has also identified the required next steps to further mature (and commercialise) the innovations proposed in ePatriot, that will be part of the next phase planned for ePatriot."