Website Orbital Eye
Three new internship opportunities are open as of today! All positions touch on the main satellite sources used by CoSMiC-EYE:
1) Synthetic Aperture Radar
3) High-resolution optical
Orbital Eye is a company based in Delft, the Netherlands. From our offices in Delft, we serve customers worldwide, in North- and Central America, Europe, and Asia. Orbital Eye was founded a little over 10 years ago and has since then been focused on developing and providing satellite-based monitoring solutions, with a focus on its flagship product pipeline monitoring solution CoSMiC-EYE.
The CoSMiC-EYE solution uses data from weather-independent Synthetic Aperture Radar (SAR) satellites, as well as multispectral, and optical satellites, to detect any activity near the assets of our customers that could damage these assets. It is mainly used to detect Third Party Activities around pipelines, such as excavation- and construction activities. Satellites scan entire countries multiple times per week, within seconds, which makes it the perfect instrument to monitor (large) pipeline networks. Our customers have the CoSMiC-EYE application installed on their devices, on which they have an overview of all activities taking place close to their assets, along with optical satellite data for each location where an activity is detected. The core of the CoSMiC-EYE technology is based on classical change detection algorithms as well as on an increasing number of Machine Learning approaches.
1) Recent R&D developments in Machine Learning have enabled the use of automatic change detection of high-resolution optical images. We are dealing with 30-50cm data acquired by different satellite providers. The first additional filtering steps have been applied to these results to decrease the number of false alarms. The goal of the internship is to further explore the results of the change detection output and develop further advanced filtering techniques to accurately pre-classify the detected activities and filter the irrelevant results. Possible research approaches may explore the use of time-series analysis, the development of pre-classification algorithms, or machine learning to solve the issues outlined above.
2) A key component in our processing chain is the pre-processing of the radar data to prepare the time series that are fed into the change detection and Machine Learning frameworks. The current processing of the radar data isn’t suitable to create time series over longer timeframes. Keeping the same statistical relations in the data over time is important when Machine Learning algorithms are trained and applied in production. This internship will look at low-level radar imagery to create stable time series data over time. A data-heavy internship where you can improve your programming skills and learn more about the infrastructure necessary to run and apply Earth Observation analysis on a global scale.
3) Multi-spectral satellites have often a lower temporal frequency compared to the radar equivalent due to clouds, however, they contain very rich information that can be used by algorithms to classify the type of detected activities. Enriching the radar data with multi-spectral information (and vice versa) would be the main objective of the internship. Therefore, offering you hands-on experience with complex Earth observation data. Combining the features of time series with different frequencies is a challenge and will require research as well as programming needs.
Each of these three internship positions can be the challenge you are looking for. Next to the exciting work, Orbital Eye offers a dynamic working environment with a young and talented team, with colleagues from many backgrounds. It’s not only work at the office, but a lot of fun activities take place as well. Think of Friday afternoon drinks, team dinners, and sports activities!
To apply for this job email your details to firstname.lastname@example.org