Private companies such as SpaceX are building large satellite constellations in low-Earth orbit (LEO) to provide global broadband internet access. For serving user requests, the satellites of such constellations fetch data from the appropriate ground station and pass it via inter-satellite links to the user’s satellite. Using well-proven CDN approaches could help to reduce the amount of network hops needed for serving common user request.
As part of the LEO-CDN student project, which took place in summer term 2021, the students developed a LEO-CDN software based on FReD and Celestial, two frameworks from our research group. The student’s software proactively migrates data between satellites to keep it close to its users. This is necessary since satellites, in contrast to ground-based servers, continuously orbit the earth with high velocity.
When driving around the city, it is not uncommon to encounter cars parking on bike lanes or similar offences. The current procedure for reporting such parking infringements is to file a manual report with the Ordnungsamt. This is time consuming and labor intensive.
As part of the weg-li student project, which took place in winter term 2020, the students developed a recommender service for the weg-li platform which simplifies the reporting process. They also developed a smartphone application for violation reporting on the go.
When evaluating new FaaS approaches for function placement, benchmarking, or routing, every group of researchers typically develops and sets up their own testbed. This makes it very difficult to reproduce results or compare them against results from other researchers.
As part of the FaaSterMetrics student project, which took place in summer term 2020, the students developed a comprehensive measurement framework for FaaS experiments. This framework supports the automatic deployment and execution of experiments and can be used in conjunction with multiple FaaS providers.
Polls show that a lack of safety and fear of accidents keeps people from using their bikes more frequently. Unfortunately, it is quite hard from a city planning perspective to get a good overview of dangerous locations since official accident statistics only cover crashes but do not provide information on near miss incidents. The SimRa project aims to collect data on such near crashes to identify when and where bicyclists are especially at risk.
There have been multiple SimRa student projects between summer term 2019 and summer term 2020. As part of these projects, the students built various visualization platform prototypes and supported the integration of the OpenBikeSensor, a Stuttgart-based hacker project that aims to measure the passing distance of overtaking vehicles through distance sensors.
Incident management includes many repetitive tasks that, in the end, still require a human in the loop for the final decision making. For making a decision, operators need a simple and intuitive process for collecting needed information on the incident, as well as a quick and easy way for initiating needed actions to resolve the incident.
As part of the ChatBots for IT Ops Incident Management student project, which took place in winter term 2018, the students developed a ChatBot applications that fully runs on AWS infrastructure. The ChatBot builds upon AWS Connect to let users interact with the chatbot by phone.
In public video surveillance, there is an inherent conflict between public safety goals and privacy needs of citizens. Generally, societies tend to decide on middleground solutions that try to balance safety and privacy goals.
In the Subject-driven Video Surveillance student project, which took place in summer term 2018, the students developed a prototype that leverages the inherent geo-distribution of fog computing to preserve privacy of citizens while still supporting camera-based digital manhunts of law enforcement agencies.
Fog computing is an emerging computing paradigm that uses processing and storage capabilities located at the edge, in the cloud, and possibly in between. Testing or running experiments in the fog, however, is hard since runtime infrastructures will typically be in use or may not yet exist.
There have been two MockFog student projects, one in summer term 2018 and one in winter term 2019, that worked on developing an approach for emulating fog infrastructures in the cloud. Furthermore, the students also studied how various data generators can be integrated into the MockFog approach for developing an automated benchmarking toolkit.