GI-Forum WS 2015/16

Date	Title	Presenter	Affiliation
20.10.2015	Institute meeting
27.10.2015	On generating spatio-temporal data	Edzer Pebesma	University Münster
	In order to gain knowledge, we (scientists) often generate new data products from primary observation data, and disseminate what we did by (i) text, e.g. in the form of a scientific publication, and (ii) publishing the input and output data and computational procedures used, e.g. an R script along with software version information. Although this may be sufficient to understand, reproduce and verify the work, it is still too difficult to assess whether two independently generated products can be compared, or whether method Y can be meaningfully applied to data set X. And how can we discover all datasets generated by procedure Z, or advertise the one we just generated? We [1] introduce a generative algebra for spatio-temporal information. Using functions on the basic types Space, Time, Quality, and Entity, we construct data generation procedures (fields, lattices, point patterns, objects, events, trajectories) that can be executed to generate actual data. Data derivation operations (the algebra) are used for generating new data types, e.g. for generating objects from fields, fields from objects, or lattices from trajectories. As opposed to data which is always discrete, data generation procedures can be continuous. They also make the data support explicit, i.e. whether values refer to points/time instances or to areas/time intervals. In contrast to the when/where/what questions usually addressed by semantics, we believe that the algebra can be used to describe the why and how questions, and as such describe data provenance. [1] joint work with Simon Scheider, Benedikt Graeler and Christoph Stasch; see http://www.meaningfulspatialstatistics.org
03.11.2015	Point Cloud Analytics	Jürgen Döllner
	In this talk, 3D point clouds, their properties, and related challenges for efficient processing and visualization of massive 3D point clouds are presented. 3D point clouds represent powerful general-purpose, discrete models of 3D objects such as in the case geospatial 3D point clouds captured by LIDAR. However, their applications go far beyond capturing reality at a given moment in time. Point cloud analytics explores and investigates how to efficiently manage large-scale, massive 3D point clouds and how to obtain actionable insights for the corresponding fields of application. Point cloud analytics will be illustrated along 'vegetation analytics', 'differential analytics' applied to 4D point clouds as well as by a use case in heavy load transportation planning. In addition, a short outline of underlying LOD and classification algorithms and data structures using GPU-acceleration is given.
10.11.2015	Geoinformatics, Open Source and Videos - A library perspective	Peter Löwe	TIB Hannover
	This presentation provides an introduction to the TIB|AV Portal for audiovisual scientific-technical information (STI) for open source - based Geoinformatics. Digital audiovisual content has become an important communication channel In Science. This content is currently hosted on publicly available commercial web portals whose functionalities do not meet the needs for reliable long term scientific preservation, access, and citation. Since the production of audiovisual content diversifies and accelerates, best practices are needed to address and solve this challenge. The TIB|AV-Portal for audiovisual STI meets the requirements to preserve such content and to provide innovative services for search and retrieval. Quality checked audiovisual content from Open Source Geoinformatics communities is constantly being acquired for the portal as a part of TIB's mission to preserve relevant content in applied computer sciences for science, industry, and the general public.
17.11.2015	The road to Paris: What is the role of Geoinformatics?	Gilberto Camara
	This talk will present an overview of the upcoming Climate Change COP-21 Conference in Paris. First, we will consider why COP-21 is different from previous climate change conference, what is stake and what are the expected aims of the conference. The talk will consider how the countries that are the major greenhouse gas polluters have promised to reduce their emissions. The final part of the talk will discuss how Brazil prepared its pledge of emission reductions, and what was the role of Geoinformatics in helping build the country's commitments.
24.11.2015	Observing the Data Universe	Tomi Kauppinen
	With the avalanche of data it becomes timely to observe the data itself to reveal interesting spatial, temporal and thematic characteristics. This is a substantial paradigm shift as we move our focus from producing more data - which automated technical sensors anyway do - to discover new law-like patterns from data. In this talk I introduce the concept of data universe and discuss via examples the potential it offers for science. The talk is organized as an interactive session with the goal of facilitating both a common understanding of the relevant concepts and supporting the emergence of a new research agenda for the years to come.
01.12.2015	Institute meeting
08.12.2015	Retrieving “invisible” information – from optical remote sensing of rainfall, birds and grasshoppers	Thomas Nauss	University Marburg
	While beginner courses on remote sensing generally focus on the detection of entities with similar size to the sensor’s spatial resolution, many interesting remote sensing applications aim in the retrieval of information about features which are magnitudes smaller and hence “invisible” to the sensor. In this talk, I will discuss three examples dealing with the retrieval of rainfall from Meteosat observations, the prediction of bird species from LiDAR data and the prediction of grasshoppers from MODIS imagery.
15.12.2015		Maria Vasardani	University Melbourne
22.12.2015 05.01.2016	No GI Forum
12.01.2016	Adaptive digital eco systems: the role of human behaviour measurement in intelligent system design	Antonio Krüger	University Saarland
	In this talk, I will discuss our recent research that aims to understand how human behaviour in digital eco systems can be measured and interpreted to inform the design and implementation of intelligent assistive systems. Today, humans are surrounded by digital eco systems, consisting of many different types of mobile and stationary devices, which consist of sensors (e.g. cameras, accelerometers) and actuators (e.g. displays). I will discuss the potential of such sensors to measure human behaviour in-situ and in realtime to infer user interests, intentions and goals and subsequently how this will lead to a novel class of intelligent systems, that primarily make use of such implicit behaviorual clues rather than relying on the currently prevailing traditional paradigm of explicit interaction through dedicated input devices such as mouse and keyboard. The talk will discuss examples from the domain of personal information management on mobile devices as well as examples from our research on instrumented environments.
19.01.2016	GEO-C: Doctoral Training Program on Enabling Open Cities	Auriol Degbelo
26.01.2016	Communication Systems for the IoT	Mesut Günesh	University Münster
The predicted number of devices connected to the Internet is around 20 - 50 billion computers for 2020. This shows that computer and ICT systems will be integrated in all parts of the society in future. The driving factor is the Internet of Things (IoT), which consists of a huge number of smart objects. A smart object is an embedded system with low computation and communication resources, which requires an efficient deployment of the available resources. In the Internet of Things all smart objects shall be able to communicate (directly) with each other, which puts strong requirements on the future communication systems. The presentation will give an introduction to the research area of the ComSys working group and present the Münster-IoT-Lab.
02.02.2016	Big Data and Data Mining in Fire Services – a first perspective	Dr. Bodo Bernsdorf, EFTAS
	After a brief description of the speaker and the company EFTAS Remote Sensing it will be described what the entire idea of Micro Rapid Mapping is based on. Programs like the International Charta “Space and major Disasters” and the Copernicus Emergency Management Service are bringing a huge amount of spatial (satellite) data into the process which are analyzed together with archive data such as aerial images and vector data such as road networks, infrastructure, etc. This concept was transferred to the needs of technical accidents like CBRN incidents in the transportation branch. In terms of investigating such situations there is the need for fire brigades to use chemical suits to protect fire fighters. This means quite intense logistics behind such a decision: Within 15 Minutes the platoon leader has to ensure the installation of the decontamination place and allocate another couple of fire fighters as rescue troop in minimum. This starts a chain of logistic supply on man and material because a troop can work only for 10 to 15 minutes on site. Afterwards another troop in other chemical suits is in charge and another rescue unit has to be in place - and so on… The idea now is to use UAV in combination with special sensors. Beside a classical RGB camera an infrared (thermal) and a hyperspectral camera will deliver pictures from the accident site. Using such sensors it is aimed to produce an emergency map with most critical information. This should be combined with classical spatial information out of municipal spatial data infrastructures such as aerial photographs, maps, boundaries of protection areas like natural or water reserve areas. But what happens? What are the needs? To be faster than the 15 minutes addressed above, it is necessary to bring the relevant data near real time to a base station for processing them. The data has to be spatially referenced and an orthophoto has to be produced. The result should be implemented as an additional layer into the SDI dataset. For this one need a quite fast radio downlink to send data while the UAV is still operating. But even then, the data tensor, the data cube produced by the different sensors is much too large for transferring all information. A data reduction has to be implemented onboard before transferring relevant pictures because the data analysis can only be fast if the information is reduced to a minimum. The photogrammetry processes are only on time if one can reduce the coverage of the accident site to at least, five, six pictures. The processing time raises quadratic with each picture coming into the process. And because the computer of the operation control car is not comparable with a computer center and due to possible missing online link one cannot use the cloud approach the calculation has to be done onsite. The BigGIS project addresses such limitations by using data mining methods to reduce the dimensions. Some examples will be described. In addition IfGI together with EFTAS started an approach to bring the idea of spatial information into the fire service education. Starting with two bachelor thesis a show case was produced by students which are used for describing the possibilities. In August 2016 there will be a first seminar at the Institute of Fire Services NRW, the central education unit for executive fire chiefs, to explain the benefit of spatial information in supporting mission control.
09.02.2016	Institute meeting