Field: Operations and R&D
Speciality: Image processing

Subject:

Vision-based navigation is an enabler for many space applications like in orbit servicing, landing on asteroids, debris removal. However the processing required for such needs shall deal with the very
limited resources of the spacecraft: algorithms have to be heavily optimized in that perspective. Besides, the need for precision remains important.

One lead to increase processing efficiency or to reach new performance levels is to rely on machine learning; the tremendous evolution in this domain (e.g. deep learning) opens new possibilities.

The training period will consist in applying some machine learning technics to visionbased algorithms developed in the department, within the Image Chain department of the Space System business line of Airbus Defence & Space.

The subject will be more precisely detailed during the interview, depending on the different studies of the image department and on the aspirations of the candidate. Depending on the planning of the internship, the contents may be adapted to be as interesting and suitable as possible.

The trainee shall have both solid image processing and machine learning backgrounds, as well as demonstrating sound computer programming skills (C/C++ or Python).

Company background:

The Space System business line of Airbus Defence & Space is the European leader in the field of optical Earth Observation systems. The company, through is history, is a pioneer of space industry, responsible for the development of the first Earth Observation space systems in Europe, starting with the SPOT family. Since this time, the company has led the major European developments in the fields, through programs such as METOP, ERS, ENVISAT, HELIOS, PLEIADES, SPOT6/7 or GAIA. This experience developed is now applied on export turn-key programs such as FORMOSAT, THEOS, ALSAT, CHILI or KazEOSat-1, involving up to sub metric resolution systems, or such as COMS, a geostationary meteorological satellite for Korea.

This evolution conveyed Airbus Defence & Space to develop a strong expertise in Image Quality, Image Processing and Image Simulation through a group of about 50 engineers in 2016, constituting the Image Chain department (TSOTU2). The Image team carries out activities in fundamental image domains such as image simulation, ground processing, image quality, in-orbit testing, embedded processing, vision-based navigation and dedicated R&D activities.

- Engineering school or Master, with specialisation in signal and image processing, machine learning or applied mathematics.

- Generic knowledge in image processing as well as numerical analysis.
- Generic knowledge or first experience with machine learning.
- C/C++, Python; Windows & Linux OS.

IMPORTANT: La data límit per sol·licitar plaça en aquest projecte és anterior als terminis habituals establerts per a sol·licitar plaça de mobilitat a l'ETSETB (aquest curs 31 de gener de 2017). Per aquest motiu cal contactar directament amb l'empresa i un cop es rebi acceptació d'aquesta, contactar amb l'ETSETB via students.mobility@etsetb.upc.edu

16/12/2016

Field: Operations and R&D
Speciality: Product software development, image processing and GUI

Subject:

Persistent observation will be soon possible from space (in geostationary orbit) and stratosphere.

Such innovative platforms will provide videos covering very large field of view: each frame will contain from 100 to 1000 mega pixels.

The Image Chain department provides, for the operational ground segment, a Catalogue of satellites images. This product should be used in the future for the referencing of these videos.

The training period will consist in several tasks:
- Developing: video metadata extraction module
- Image processing: reformatting videos to be readable in internet browser
- GUI: Search and visualise videos in the Image Chain Catalogue.

The subject will be more precisely detailed during the interview, depending on the different studies of the image department and on the aspirations of the candidate. Depending on the planning of the internship, the contents may be adapted to be as interesting as possible.

The trainee shall have both computer programming skills (JAVA EE, GWT, JavaScript, Servlet Java, XML, C/C++, and Python) and an image processing background. We use "agile" methodologies and work in a continuous integration and automated validation environment.

Company background:

The Space System business line of Airbus Defence & Space is the European leader in the field of optical Earth Observation systems. The company, through is history, is a pioneer of space industry, responsible for the development of the first Earth Observation space systems in Europe, starting with the SPOT family. Since this time, the company has led the major European developments in the fields, through programs such as METOP, ERS, ENVISAT, HELIOS, PLEIADES, SPOT6/7 or GAIA. This experience developed is now applied on export turn-key programs such as FORMOSAT, THEOS, ALSAT, CHILI or KazEOSat-1, involving up to sub metric resolution systems, or such as COMS, a geostationary meteorological satellite for Korea.

This evolution conveyed Airbus Defence & Space to develop a strong expertise in Image Quality, Image Processing and Image Simulation through a group of about 50 engineers in 2016, constituting the Image Chain department (TSOTU2). The Image team carries out activities in fundamental image domains such as image simulation, ground processing, image quality, in-orbit testing, embedded processing, vision-based navigation and dedicated R&D activities.

- Engineering school or Master, with specialisation in signal and image processing, or applied mathematics.

- JAVA EE, GWT, JavaScript, Servlet Java, XML, C/C++, and Python.
- Generic knowledge in image processing.
- Windows & Linux OS.

Field: Operations and R&D
Speciality: Image processing

Subject:

Image performance is the key components of the acquisition chain for observation missions and planetary exploration missions. Needs in processing are of various types: restoration, registration, image fusion, tracking, navigation, image simulation…

The training period will consist in developing and testing new algorithms in one of the above-cited domain within the Space System business line of Airbus Defence & Space. The subject will be more precisely defined during the interview, depending first, on the different studies of the image department and second, on the aspirations of the candidate.

The subjects already identified this year are in the following themes:
- Colour fusion (demosaicing and Pan Sharpening to compute high resolution colour image from low res colour and high res B&W images)
- Very precise imaging geometry for high resolution satellite

Depending on the planning of the internship, the contents may be adapted to be as interesting as possible.

The trainee shall have both a solid image processing background and computer programming skills (Matlab and/or C/C++).

Company background:

The Space System business line of Airbus Defence & Space is the European leader in the field of optical Earth Observation systems. The company, through is history, is a pioneer of space industry, responsible for the development of the first Earth Observation space systems in Europe, starting with the SPOT family. Since this time, the company has led the major European developments in the fields, through programs such as METOP, ERS, ENVISAT, HELIOS, PLEIADES, SPOT6/7 or GAIA. This experience developed is now applied on export turn-key programs such as FORMOSAT, THEOS, ALSAT, CHILI or KazEOSat-1, involving up to sub metric resolution systems, or such as COMS, a geostationary meteorological satellite for Korea.

This evolution conveyed Airbus Defence & Space to develop a strong expertise in Image Quality, Image Processing and Image Simulation through a group of about 50 engineers in 2016, constituting the Image Chain department (TSOTU2). The Image team carries out activities in fundamental image domains such as image simulation, ground processing, image quality, in-orbit testing, embedded processing, vision-based navigation and dedicated R&D activities.

- Engineering school or Master, with specialisation in signal and image processing, or applied mathematics.

- Generic knowledge in image processing as well as numerical analysis.
- Matlab or C/C++; Windows & Linux OS.

Field: Processing Development and Studies
Speciality: Image processing

Subject:

Image production subsystems require light and fast image processing modules. Space Systems have developed a family of image processing modules for image data acquisition, radiometric
and geometric corrections.

The objective of the internship is to implement operational image processing algorithms to extend or replace existing operational functions in our operational image processing modules, keeping a
high technical quality/performance level (accuracy, timeliness), within an efficient process in terms of development and maintenance.

The technical area may concern radiometric or geometric correction processing, or synchro-decommutation / inventory processing, for still or video optical images.

Operational image processing software will be developed in C++ and Python languages. Matlab might be also used in the course of the activities.

A particular emphasis shall be put on code implementation optimisation and reusability.

Company background:

The Space System business line of Airbus Defence & Space is the European leader in the field of optical Earth Observation systems. The company, through is history, is a pioneer of space industry, responsible for the development of the first Earth Observation space systems in Europe, starting with the SPOT family. Since this time, the company has led the major European developments in the fields, through programs such as METOP, ERS, ENVISAT, HELIOS, PLEIADES, SPOT6/7 or GAIA. This experience developed is now applied on export turn-key programs such as FORMOSAT, THEOS, ALSAT, CHILI or KazEOSat-1, involving up to sub metric resolution systems, or such as COMS, a geostationary meteorological satellite for Korea.

This evolution conveyed Airbus Defence & Space to develop a strong expertise in Image Quality, Image Processing and Image Simulation through a group of about 50 engineers in 2016, constituting the Image Chain department (TSOTU2). The Image team carries out activities in fundamental image domains such as image simulation, ground processing, image quality, in-orbit testing, embedded processing, vision-based navigation and dedicated R&D activities.

- Engineering school or Master, with specialisation in signal and image processing, or applied mathematics and software development.

- Generic knowledge in signal and image processing as well as numerical analysis.
- C++, Python, Matlab.

Field: Operations and R&D
Speciality: Satellite guidance and mission optimisation

Subject:

Earth observation satellites are acquiring images in difficult geometrical conditions: high orbiting satellite speed, Earth projection effects, optical distortion, and so on. In order to optimize the image quality, the satellite orientation is controlled on its 3-axes. Such control is called “guidance” and is the result of a complex optimization with respect to geometrical criteria. This computation may lead to non-linear ground projection trajectory or non-predictable acquisition duration.

While guidance is computed with precision for satellite control, there is a need for fast guidance computation. Mission department of Airbus DS is using guidance method to compute the plan of a satellite for several requests. Fast guidance allows the mission chain to develop new algorithms evaluating multiple guidance solution to select the best solution in multi-satellites context.

Therefore the internship consists, on one hand, in the design of the fast-guidance method based on a neural-network method and deep learning techniques. On the other hand, it consists in the integration of this procedure in mission chain algorithms and to measure the error of this fast-prediction method. This internship is a unique opportunity for the trainee to discover, learn and work in two complementary departments: mission and image chains.

Company background:

The Space System business line of Airbus Defence & Space is the European leader in the field of optical Earth Observation systems. The company, through is history, is a pioneer of space industry, responsible for the development of the first Earth Observation space systems in Europe, starting with the SPOT family. Since this time, the company has led the major European developments in the fields, through programs such as METOP, ERS, ENVISAT, HELIOS, PLEIADES or SPOT6. This
experience developed is now applied on export turn-key programs such as FORMOSAT, THEOS, ALSAT, CHILI or KazEOSat-1, involving up to sub metric resolution systems, or such as COMS, a geostationary meteorological satellite for Korea.

This evolution conveyed Airbus Defence & Space to develop a strong expertise in Image Quality, Image Processing and Image Simulation through a group of about 50 engineers in 2016, constituting the Image Chain department (TSOTU2). The Image team carries out activities in fundamental image domains such as image simulation, ground processing, image quality, in-orbit testing, embedded processing, visionbased navigation and dedicated R&D activities.

- Generic knowledge in image processing as well as numerical analysis
- Matlab or C/C++; Windows & Linux OS
- CNN, cloud computing

Desired education:
- Engineering school or Master, with specialisation in signal and image processing, or applied mathematics.

IMPORTANT: Contactar directament amb l'empresa i un cop es rebi acceptació d'aquesta, contactar amb l'ETSETB via students.mobility@etsetb.upc.edu

This opportunity is a one-year internship at BMW Group Headquarters in Munich, in the areas where the next exciting development in the automative industry happens, e.g. digitalization, artificial intelligence, autonomous driving.We are looking for 15 young people with enthusiasm for starting their careers, with initiative and passion about the Brand.

• Bachelor degree in IT or Engineering or related to STEM.
• B2 English Level.
• Enthusiasm in the automotive industry and the willingness for further qualification in future competencies.
• Passion about mobility solutions for tomorrow.
• Ready to live a great professional and personal experience.

Subject:

Satellite remote sensing imagery represents an attractive data source to monitor large regions with frequent updates. In this context, the operational production of accurate land cover maps plays an important role in global-scale environmental
assessments and becomes crucial for a wide range of research domains. New
earth observation missions such as Sentinel provide images with high spatial and
temporal resolution. Accordingly, new image classification methods for the
generation of reliable land cover maps are needed.

Hosting Laboratory:
The CESBIO laboratory (Centre d'Etudes Spatiales de la BIOsphère) in Toulouse,
France. CESBIO aims to develop knowledge on continental biosphere dynamics and
functioning at various temporal and spatial scales. The Cesbio is a fourth
partnership laboratory: 1) CNES (the French Space Agency), 2) CNRS (National
Center Scientific Research), 3) UPS (Paul Sabatier University), and 4) IRD
(Institute of Research for Development).

Subject: The student will work in topics related to deep learning and multimedia data analytics. S/he will be assigned into a team of researchers and actively participate in the daily routine of the institution. The exact topic of research will be discussed and decided after the selection from the school, based on the interest and priorities of DFKI and the student.

Research center: The German Research Center for Artificial Intelligence, with sites in Kaiserslautern, Saarbrücken, Bremen (with an associated branch in Osnabrück) and a project office in Berlin, is the leading German research institute in the field of innovative software technology. In the international scientific community, DFKI ranks among the most recognized "Centers of Excellence" and currently is the biggest research center worldwide in the area of Artificial Intelligence and its application in terms of number of employees and the volume of external funds.

Bachelor: Around 5 months, can be discussed.

Master students: Between 6-12 months, can be discussed.

DFKI is searching for excellent students with a background on deep learning and its related software frameworks. English at a professional level is required, as this is the common work language in the center.

In order to prepare the intersnhip, students are encouraged to enrol in the “Deep Learning for Computer Vision” and “Deep learning for Speech and Language” seminars, as well as participate in the Computer Vision Reading Group with associated ECTS for master MET students. Please contact professor Xavier Giró for more details on these opportunities (xavier.giro@upc.edu).

Subject: The Optical Communications Group investigates new technologies to improve performance of free-space optical links between ground stations, aircrafts, satellites, etc. Optical communications systems nowadays aim towards higher spectral efficiency, increased data rate and improved sensitivity of the receiver. Widely used on-off keying modulation, together with direct detection, offers a relatively low system implementation complexity. Coherent modulation and detection techniques have also been studied and implemented. They offer improved sensitivity, but require an optical PLL or post-processing techniques in the receiver.
In the framework of this thesis the student will have to study, simulate, implement and experimentally evaluate performance of a given PSK modulation and detection system in terms of the achievable sensitivity and BER vs. OSNR, required linewidth and optical power of the local oscillator, etc. During the thesis the student will focus on the technical challenges of the system implementation and demonstration under laboratory as well as outdoor conditions, taking into account various noise sources, HW and SW limitations, etc.

Tasks:
The subject offers several internship and Master Thesis topics like:

- Proposing an improved detection technique taking into account the system and channel model.

- Design, setup and verification of the complete optical link in the laboratory and outdoors.

- Performance analysis of individual modulation and detection techniques under given conditions.

- Derivation of an improved atmospheric channel model and its implementation into a simulation tool.

It is an excellent opportunity to work in a cutting edge topic, gain experience on real optical hardware, and improve your engineering skills.

Research center: The German Aerospace Center (DLR) is Germany´s national research center for aeronautics and space and Germany´s space agency. Its extensive research and development work in Aeronautics, Space, Transportation and Energy is integrated into national and international cooperative ventures. Approximately 5,500 people are employed in DLR´s 28 institutes and facilities at eight locations in Germany.

Subject: The optical communication groups of DLR’s Institute of Communications and Navigation develop communication systems for optical links between satellites and optical ground stations. The trend goes to the usage of fiber-based multiplexing technologies which enable data rates up to 1Tbit/s. For this scenario, an optical communication system with a precise fine pointing and fiber coupling system is developed.

Tasks:
This fiber coupling system requires precise alignment and adjustment of the optomechanical components and sensors. This will be achieved with an automated calibration setup. In the thesis, the following points will be included:
- Development of the calibration setup and calibration methodology.
- Implementation on a DSP.
- Practical verification and optimization.
- Conduction of field-tests.

Research center: The German Aerospace Center (DLR) is Germany´s national research center for aeronautics and space and Germany´s space agency. Its extensive research and development work in Aeronautics, Space, Transportation and Energy is integrated into national and international cooperative ventures. Approximately 5,500 people are employed in DLR´s 28 institutes and facilities at eight locations in Germany.

Subject: The optical communication groups of DLR’s Institute of Communications and Navigation develops experimental laser communication systems for optical data links between airborne carriers such as airplanes or satellites and optical ground stations. To achieve a reliable data reception also under difficult atmospheric situations, optical sensors with a large area (diameter ≈ 1mm) should be used. The link distances of the systems are typically in the range of several 10 to hundreds of km and the receivers have to operate with very low input power. Hence, optical detectors and electrical circuits have to be highly sensitive. The amplifier circuits have to provide a high gain bandwidth product combined with a very low noise.

Tasks:
This thesis includes the development of the analog and mixed signal circuits for the described optical receiver. Furthermore it includes:
- Circuit analysis based on SPICE.
- Circuit design with Altium Designer.
- Layout development and assembly.
- Lab verification, optimization and test.

Research center: The German Aerospace Center (DLR) is Germany´s national research center for aeronautics and space and Germany´s space agency. Its extensive research and development work in Aeronautics, Space, Transportation and Energy is integrated into national and international cooperative ventures. Approximately 5,500 people are employed in DLR´s 28 institutes and facilities at eight locations in Germany.

Subject: The subject of the work will be the extension of SDN (Software Defined Networking) architectures to satellite payloads, in order to implement efficient routing, load balancing, and quality of service strategies in LEO megaconstellations. A full SDN network will be assumed and the advantages coming from the application of SDN onboard satellite will be assessed theoretically and/or simulatively depending on the background of the candidate. Alternatively to LEO megaconstellations, the case of hierarchical configurations composed of LEO and GEO tiers can be also taken into consideration.
Motivation for this study comes from the emergence of new very large constellations (OneWeb, O3b, etc.) and similar concepts also proposed by Facebook and Google, which require advanced routing and overall networking mecahnisms. Possible answer to some of these challenges is offered by the SDN paradigm, which now out since about 5 years and has received a lot of consensus in the terrestrial community.

Tasks:
- State of the art of routing solution in LEO systems.
- SDN architectures and concepts already proposed for use in satellite domains.
- Design of novel routing schemes and related validation through simulation campaigns.

Research center: The German Aerospace Center (DLR) is Germany´s national research center for aeronautics and space and Germany´s space agency. Its extensive research and development work in Aeronautics, Space, Transportation and Energy is integrated into national and international cooperative ventures. Approximately 5,500 people are employed in DLR´s 28 institutes and facilities at eight locations in Germany.

Candidate must be able to take high level architectural design input and implement that design in hardware.  
Must proficient in both analog and digital design.
Should be able to calculate system power requirements and implement a power solution that meets those requirements.
Candidate must have fabrication skills, including printed circuit board assembly, including surface part mounts, as well as cabling fabrication, and chassis assembly. 

Must have proven testing and troubleshooting skills with the use of lab tools like signal generators and oscilloscopes.
An understanding of ultra high vacuum systems would be a plus.
Familiarity with Labview or Matlab and willingness to develop further proficiency.
Be proficient in the use of CAD tools, e.g. ConceptHDL schematic capture, Orcad Capture, or similar schematic tool.

The ideal candidate will be able to generate net lists, resolve errors in report generation, and covert schematics from one tool to the other.  
Experience in PCB layout and tools such as Allegro is a plus.
Should have a familiarity with major computer interfaces such as PCI, PCLe, I2C, SPI, Ethernet, DRAM, SRAM, etc.

For the successful candidates, ECTalent provides and pays for the following benefits:

•    US visa application processing

•    Air travel to/from the United States

•    Free housing in comfortable, fully-furnished apartments with all utilities paid.

•    Competitive intern salary ($2000 USD per month gross)

•    Health benefits, including medical & dental insurance

•    Onboarding orientation to help you get going in the United States quickly

•    Cultural program to introduce you to US culture and the best California has to offer

•    Career coaching

•    Networking and social events

Context:
Image and Pervasive Access Laboratory is a reference laboratory on Assistive technology for aging people in Singapore and nowadays we are enrolled in 2 European projects. We target several goals in our research. The first goal is to develop and deploy continuous and unobtrusive Ambient Assistive Living platforms in real scenarios. The aim is to recognize aging Activities of the Daily Living through semantic and rule-based reasoning with various level of abstraction. The second goal is to move towards distant and large-scale deployments of AAL platforms. Therefore, we built UBISmart, an agile framework that has the advantages of using a web platform to manage a large-scale deployment; the versatility and rapidity that this solution offers in a plug & play approach; and facilitates the active participation of end-users. UBISmart is a generic, scalable and intuitive cloud Web-Based AAL framework, which gathers sensitive information from the spaces through simple devices and sensors. This AAL platform converts any environment into a smart space in five (5) minutes to assess dependent people at their homes. Finally, we are now moving from controlled spaces to wide and open spaces in a SmartCity vision.

Description:
In the frame of SmartHomes, we need to make sense of the collected data within a given a human context to improve the interaction with end-users. The proposed approach processes the collected data in real time and computes the human condition values to personalize the service delivery.
We propose a computer science subject with the following description:
• The intern will study the computation of human behaviours from ambient and unobtrusive data.
• The intern will be in charge of developing algorithms to clean heterogeneous raw data.
• The intern will be in charge of developing algorithms to compute and quantify human behaviours.
• The intern will integrate the proposed algorithm in a cloud based web application.
• The intern will work the inner functionalities of our web platform.

Keywords:
Ambient Assisted Living, REST, Javascript, Semantic Web, Data Mining.

Context:
Image and Pervasive Access Laboratory is a reference laboratory on Assistive technology for aging people in Singapore and nowadays we are enrolled in 2 European projects. We target several goals in our research. The first goal is to develop and deploy continuous and unobtrusive Ambient Assistive Living platforms in real scenarios. The aim is to recognize aging Activities of the Daily Living through semantic and rule-based reasoning with various level of abstraction. The second goal is to move towards distant and large-scale deployments of AAL platforms. Therefore, we built UBISmart, an agile framework that has the advantages of using a web platform to manage a large-scale deployment; the versatility and rapidity that this solution offers in a plug & play approach; and facilitates the active participation of end-users. UBISmart is a generic, scalable and intuitive cloud Web-Based AAL framework, which gathers sensitive information from the spaces through simple devices and sensors. This AAL platform converts any environment into a smart space in five (5) minutes to assess dependent people at their homes. Finally, we are now moving from controlled spaces to wide and open spaces in a SmartCity vision.

Description:
In this context assistive technologies need to improve the user experience to maintain and reinforce the social link between actors. We aim to provide simple and integrated applications combining self-coded features with wearables communications, products API and social networks.
We propose a computer science subject with the following description:
• The intern will be in charge of developing algorithms to connect industrial wearables to our cloud platform.
• The intern will be in charge of developing mobile apps (Android, Ionic) to gather embedded sensor data.
• The intern will be in charge of resolving Server-client communications issues.
• The intern will work on the Integration of products API into web platforms and mobile Apps.
• The intern will be in charge of developing state of the art interfaces to improve the user experience.

Keywords:
Ambient Assisted Living, Smart Devices, REST, Mobile Apps, Javascript.

Context:
Image and Pervasive Access Laboratory is a reference laboratory on Assistive technology for aging people in Singapore and nowadays we are enrolled in 2 European projects. We target several goals in our research. The first goal is to develop and deploy continuous and unobtrusive Ambient Assistive Living platforms in real scenarios. The aim is to recognize aging Activities of the Daily Living through semantic and rule-based reasoning with various level of abstraction. The second goal is to move towards distant and large scale deployments of AAL platforms. Therefore, we built UBISmart, an agile framework that has the advantages of using a web platform to manage a large-scale deployment; the versatility and rapidity that this solution offers in a plug & play approach; and facilitates the active participation of end-users. UBISmart is a generic, scalable and intuitive cloud Web-Based AAL framework, which gathers sensitive information from the spaces through simple devices and sensors. This AAL platform converts any environment into a smart space in five (5) minutes to assess dependent people at their homes. Finally, we are now moving from controlled spaces to wide and open spaces in a SmartCity vision.

Description:
In the context of SmartCities, we need to keep collecting data once the person leave the controlled space and move around the city. We aim to merge new information coming from wearables and social networks data with the profile and collected data with our system to deliver improved services to end-users.
We propose a computer science subject with the following description:
• The intern will be in charge of developing algorithms to connect industrial wearables to our cloud platform UbiSMART.
• The intern will be in charge of developing mobile apps (Android, Ionic) to gather embedded sensor data.
• The intern will be in charge of making sense of the collected raw data to improve the assessment.
• The intern will work with multiples teams (Assistive technology, Big Data) to improve his algorithms and feed them with social and human data.
• The intern will work on the Integration of products API into web platforms and mobile Apps.

Keywords:
Ambient Assisted Living, Smart Devices, MobileApp development, REST, Javascript, Semantic Web, Big Data.

Presentation of the project
Segmentation of objects into an image is still a challenging problem in
general and particularly in biomedical imaging. Biomedical images are
usually complex and noisy, so classical segmentation methods based on
thresholding or edge detection are not adapted. A possible approach is to
use super-pixels (see figure 1), where pixels of similar content (color or
intensity) are clustered together. This approach hence helps to reduce the
complexity of the image to thousands of super-pixels region instead of
millions of individual pixels.

Expected deliverables
The project of this internship is not to implement super-pixels methods but
to study how to use super-pixels for segmentation of 2D and 3D data in
biomedical images. The super-pixels image can be apprehended as a graph
connecting the different super-pixels regions. The project will then consist
in applying graph functionalities and properties to super-pixels. Various
other ideas may also be implemented like merging of two regions, regions
dilation, erosion, ...

Keywords
Segmentation, graphs, super-pixels, 2D, 3D, biomedical imaging

Our Research:

Many city-dwelling elderly people can be greatly affected after a minor change in their living or health conditions. Mild Cognitive Impairment (MCI), early dementia and frailty are among the most common risks with deep consequences on the elderly people's and caregivers' quality of life. Through the new wave of Information and Communication Technologies (ICT), Internet of Things (IoT) and Smart City systems, it is now possible to help individuals capture and make use of their personal data in a way that will help them maintain their independence for a longer period.
The IPAL research team and its partners aim to create an innovative service based on:

• ICT-enhanced early detection of risk related to frailty
• ICT-enhanced interventions that can help the elderly population to improve their daily life and also promote positive behavioral changes

Through real-life pilot sites in Singapore in collaboration with local stakeholders, the IPAL research team explores how data on individual behaviors captured through indoor and outdoor sensors could be used for the observation and detection of the following parameters:

• Activity of Daily Living (ADL): nutrition, hygiene, sleep activity
• Mobility: physical activity, going-out frequency and going-out length
• Cognition: forgetfulness, early signs of cognitive decline
• Socialization: senior activity centre visits, activities attended, other places of interests visits

Our focus is to use sensing technologies installed in the elderly’s environment (indoor to outdoor) to monitor and detect their activities of daily living. Sensor data that is collected will then be analysed to identify relevant behaviors of individuals, and to detect behavioral changes that can be correlated with risks of MCI/frailty. The appropriate ICT based interventions (e.g. data visualization and alerts to caregivers) will then be applied to mitigate these risks.
The objective of this internship is to make use of this personal data, and create an interface (front-end) for the users (elders themselves, volunteers, relatives, doctors…) to observe the data in their own user experience. For example, the elders will need a really simple interface to check their daily activities performed (possibility of some gamification), the relatives may wish to observe what the elderly people are doing at the moment and some trends of activities performed, while the volunteers will need to detect a risk and potentially interact with the elderly to help in any matter.

About the Internship:

We are looking for a UI Developer to build new user interface features and create an extraordinary web experience for users, volunteers and relatives.

Responsibilities:

• Understand the usages for the platform at each level (elders, volunteers, family)
• Define the interface specifications for each scenario
• Propose an adequate and convenient user experience for each interface
• Build front end components using HTML, CSS (bootstrap), JavaScript
  Design and implement user interface components with right consideration for usability, security, performance and customizability
• Drive consistent end-to-end user experience

Keywords:
Ambient Assisted Living, Social sciences, Public health, Behavioral, Digital health, User experience, Dementia assistance, Real life deployment, Dynamic and adaptable systems, Context aware services

Applicant's profile:

•  Knowledge of command line tools, shell, Git or versioning tools (good practices).
•  Knowledge of web applications and web services.
•  Awareness in Digital Health (ICT services for elderly).
•  Knowledge of mobile apps development (JS frameworks, native iOS/Android not necessary).
•  Systematic approach.
•  Proactiveness and independence.
•  Excellent command of English, Chinese (mandarin or dialects) proficiency is a plus.

Context and Objectives:
In the context of the European Project City4Age1, our research team and its partners aim at creating an innovative service for ageing people based on:
- early detection of risk related to frailty.
- enhanced interventions that can help the elderly population improve their daily life and also promote positive behaviour change.

The work is realized through real-life worldwide pilot sites (in Montpellier, Singapore, Athens, Birmingham, Lecce and Madrid) and in close collaboration with IPAL lab in Singapore and DOMUS research lab in Sherbrooke, Canada. We explore how data on individual behaviours captured through indoor and outdoor sensors could be used for the observation and detection of the following parameters:
- Activity of Daily Living (ADL): nutrition, hygiene, sleep activity.
- Mobility: physical activity, going-out frequency and going-out length.
- Cognition: forgetfulness, early signs of cognitive decline.
- Socialization: activities attended, visits in other places of interest.

Therefore, we are working on a ambient assistive platform (i.e., Ubismart) that uses Internet of Things technology to ensure assistive services continuity, indoor (houses) and outdoor (cities). The platform is based on ubiquitous technologies (e.g., proximity, presence and contact sensors), wearable technologies (e.g., smartphone, smartwatch), and other outsourced services (e.g., weather) in order to understand a situation (of an ageing person), detect a problem, and intervene if necessary. The platform also allows to create intelligent spaces where sensors are easily and dynamically deployed, and reasoning algorithms are put in place (using semantic web).

Our research team is looking for an intern to advance its research on ambient assisted living for elderly people. The objectives of the internship are “Integrating new indoor technologies (e.g., humidity, pressure, acceleration and temperature sensors) and new communication protocols (e.g., Bluetooth Low Energy, Z-Wave and EnOcean) in our ambient assistive living platform”.

Description:
The work to be made involves the intern to:
- integrate new indoor technologies (e.g., humidity, pressure, acceleration and temperature sensors) in our ambient assistive living platform
- integrate new communication protocols (e.g., Bluetooth Low Energy, Z-Wave and EnOcean) in our ambient assistive living platform
- integrate new indoor technologies' events in the ambient assistive living platform reasoning and decision making process
- participate in real-life deployments in individual houses in Montpellier

Keywords:
Ambient Assisted Living, Public Health, Digital Health, Real life deployment, Sensors
1 www.city4ageproject.eu
2 in Montpellier, Singapore, Athens, Birmingham, Lecce and Madrid
3 www.ipal.cnrs.fr
4 domus.usherbrooke.ca

About LIRMM:
The Montpellier Laboratory of Informatics, Robotics and Microelectronics LIRMM) is attached to Montpellier University and the French National Center for Scientific Research (CNRS). Its activity is developed within three scientific research departments (Informatics, Microelectronics, Robotics), 19 team-projects and several common services.

The internship is based in Montpellier.

Montpellier Laboratory of Informatics, Robotics and Microelectronics LIRMM

Context and Objectives:
In the context of the European Project City4Age (www.city4ageproject.eu), our research team and its partners aim at creating an innovative service for ageing people based on:
- early detection of risk related to frailty.
- enhanced interventions that can help the elderly population to improve their daily life and also promote positive behaviour change.

The work is realized through real-life worldwide pilot sites (in Montpellier, Singapore, Athens, Birmingham, Lecce and Madrid) and in close collaboration with IPAL lab in Singapore and DOMUS research lab in Sherbrooke, Canada. We explore how data on individual behaviours captured through indoor and outdoor sensors could be used for the observation and detection of the following parameters:
- Activity of Daily Living (ADL): nutrition, hygiene, sleep activity.
- Mobility: physical activity, going-out frequency and going-out length.
- Cognition: forgetfulness, early signs of cognitive decline.
- Socialization: activities attended, visits in other places of interest.

Therefore, we are working on a ambient assistive platform (i.e., Ubismart) that uses Internet of Things technology to ensure assistive services continuity, indoor (houses) and outdoor (cities). The platform is based on ubiquitous technologies (e.g., proximity, presence and contact sensors), wearable technologies (e.g., smartphone, smartwatch), and other outsourced services (e.g., weather) in order to understand a situation (of an ageing person), detect a problem, and intervene if necessary. The platform also allows to create intelligent spaces where sensors are easily and dynamically deployed, and reasoning algorithms are put in place (using semantic web).

Our research team is looking for an intern to advance its research on ambient assisted living for elderly people. The objectives of the internship are “Integrating new outdoor technologies (e.g., smart phone and beacons) in our ambient assistive living platform”.

Description:
The work to be made involves the intern to:
- integrate new outdoor technologies (e.g., smart phone and beacons) in our ambient assistive living platform.
- integrate smart phones' and beacons' events in the ambient assistive living platform reasoning and decision making process.
- the work will be validated throw a real-life deployments in individual houses in Montpellier.

Keywords:
Ambient Assisted Living, Public Health, Digital Health, Real life deployment, Sensors, Smart Phone, Beacons

About LIRMM:
The Montpellier Laboratory of Informatics, Robotics and Microelectronics LIRMM) is attached to Montpellier University and the French National Center for Scientific Research (CNRS). Its activity is developed within three scientific research departments (Informatics, Microelectronics, Robotics), 19 team-projects and several common services.

The internship is based in Montpellier.

Montpellier Laboratory of Informatics, Robotics and Microelectronics LIRMM

Description
Lumileds R&D is searching for graduate and undergraduate students willing to contribute to the solid state lighting industry. The student/s will work in a high tech industry facility together with a professional research team of electrical engineers and physicists.
The proposed area of investigation deals with numerical methods and computational techniques to model the processes of light extraction in power LED (Light-Emitting Diodes) devices. An optics model of thin film micro-structures ought to be developed, tested, and experimentally validated. The student will have the opportunity to closely work with a team working on advanced multiphysics LED models to guide us in the understanding of a number of exciting fundamental physical mechanisms.

Context
Solid state lighting technologies are among the most prominent innovations influencing the way in which we improve our future by reducing energy consumption. So much so that the current 20% of the world’s electricity consumed by lighting can potentially be reduced to 4% with the full-scale adoption of LEDs.

Description
Lumileds R&D is searching for graduate and undergraduate students willing to contribute to the solid state lighting industry. The student/s will work in a high tech industry facility together with a professional research team of electrical engineers and physicists.
The proposed area of investigation deals with numerical methods and computational techniques to model the processes of photo-conversion and light scattering in power LED (Light-Emitting Diodes) emitters. An optics model of thin film micro-structures ought to be developed, tested, and experimentally validated. The student will have the opportunity to closely work with a team working on advanced multiphysics LED models to guide us in the understanding of a number of exciting fundamental physical mechanisms.

Context
Solid state lighting technologies are among the most prominent innovations influencing the way in which we improve our future by reducing energy consumption. So much so that the current 20% of the world’s electricity consumed by lighting can potentially be reduced to 4% with the full-scale adoption of LEDs.

Lumileds R&D is searching for graduate and undergraduate students willing to contribute to the solid state lighting industry. The student/s will work in a high tech industry facility together with a professional research team of electrical engineers and physicists. The specific research project they are asked to participate in relates to fundamental studies on optical properties of InGaN multi-quantum well (MQWs) for high brightness LEDs.

Optimizing spontaneous emission of light in InGaN MQWs requires deep knowledge of a number of fundamental quantum mechanical processes of low dimensional semiconductor systems, some of which comprise energy band formation in crystals, exciton dynamics, carrier recombination, quantum confinement and spin-orbit coupling. These mechanisms can be studied with the aid of widely extended tools such as those based on Fermi’s golden rule, k.p theory, Schrodinger equation and the effective mass approximation. Combined with efficient computational methods and calibration techniques, the resulting models can accurately quantify relevant performance characteristics, including polarization properties, lifetimes, absorption and emission spectra.

Responsibilities:

- Support research team on the theoretical analysis of intraband transitions in InGaN QW superlattices.

- Develop and test quantum mechanical models of InGaN based MQWs.

- Model calibration

- Data processing

Field of photonics, electronics, electrical and telecommunication engineering.

Skills: English, scientific programming, photonics, modeling.

Lumileds R&D is searching for graduate and undergraduate students willing to contribute to the solid state lighting industry. The student/s will work in a high tech industry facility together with a professional research team of electrical engineers and physicists. The specific research project they are asked to participate in relates to plasmonics for Solid State Lighting applications with particular focus on the coupling of plasmonics to luminescent Quantum Dots (QDs) and Quantum Wells (QWs).

Investigations on light-matter interactions have revealed a number of strong resonant coupling regimes in schemes involving quasi-particles such as surface plasmons, excitons, polaritons and combinations thereof. Most notably, some of these regimes are achieved by means of surface lattice arrangements of metallic nanostructure arrays (or nanoantenna arrays) wherein the hybridization between molecular excitons and surface lattice resonances has been observed. The properties of these resonant coupling regimes fall at the heart of quantum condensed state of light and matter. Of particular interest are those properties that enable the development of more efficient and highly tunable nanoscale light sources by introducing the ability to control key features such as photon energy, spectral width, directionality and polarization. The fundamental platform to establish this conveniently comprises the coupling interaction between pump light, QDs and metallic nanostructure arrays. Demonstrations have shown that emitted light from the QDs can be enhanced and controlled by structuring the metal to sustain surface plasmons that resonate with the emission. Spatial positioning is critical due to the deep sub-wavelength scale confinement of the surface plasmon resonant fields. This challenge has been tackled by the use of thin film luminescent layers of QDs deposited on substrates with metal nanoantennas supporting surface lattice resonances.

Responsibilities:

- Support research team on the theoretical analysis of plasmonics and metamaterials for specific device architectures.

- Develop computational models.

- Material characterization.

- Data processing.

Field of photonics, electronics, electrical and telecommunication engineering.

Skills: English, scientific programming, optics, photonics, modeling.

The subjects change over time and are also based on the background of the student, but will be related to state-of-the-art Speech Recognition Development.

Likely topics in Summer 2017 are:
- Robust recognition of spoken ‘wake-up’ keywords with DNN’s.
- Multi-microphone input to DNN / LSTM acoustic models.
- Robust recognition with DNN / LSTM acoustic models (e.g. training with clean audio, recognition with far-field recordings).
- Speaker diarization on multi-microphone recordings.

- Generally good programming skills in C/C++ or java – possibly Python.
- Knowledge about statistics and speech recognition (e.g. speech technologies course).
- Possibly advanced machine learning techniques or other exposure to ASR technology can work instead of speech technologies course.

User association in wireless networks has historically been done on the basis of single connectivity, i.e. the user equipment (UE) being connected to a single serving access point (AP). Such a design was quite efficient for conventional homogeneous networks with mostly voice-centric applications. However, as networks become more heterogeneous and services become diverse with different performance requirements, connecting to a single AP has proven to be quite inefficient. Connecting to more than one AP simultaneously provides access to increased resources, diversity and the opportunity to improve the load-balance in the network. Dual connectivity, an LTE release-12 feature for small cell enhancement, is a first step in this direction. Other examples of multi-connectivity include soft handover in CDMA systems, Licensed Assisted Access (LAA) in LTE Release-14 and Downlink Uplink Decoupling (DUDe).

Multi-Connectivity is foreseen to play a pivotal role in future 5G systems. The objective of this study is to perform an in-depth study of Multi-Connectivity in 5G. The study will include:

- Performance evaluation of Multi-Connectivity in 5G with ideal conditions.
- Impact of practical constraints like back haul limitation, CSI inaccuracy etc. on the performance of Multi-Connectivity in 5G.
- Control channel aspects for Multi-Connectivity in 5G.
- RRM and scheduling algorithms for Multi-Connectivity in 5G.

For many years the mobile communication industry has mainly focused on mobile broadband services when designing solutions for wireless network deployments in both licensed (e.g. LTE) and unlicensed (e.g. WiFi) spectrum. The Multefire Alliance, which was recently formed with the target of bringing LTE-based technology in unlicensed spectrum, has also primarily targeted mobile broadband applications in the first release of their specifications (November 2016). However, the Internet-of-Things (IoT) has gained more and more interest in the recent years. One of the most interesting IoT applications is massive deployments of low-complexity IoT devices in industrial environments, partially enabling what is known as Industry 4.0. Practical examples of these IoT applications are prediction and prevention of the wearing down and failure of machines and equipment, enabling positioning, tracking, monitoring and control of various plant assets, as well as augmented reality applications for plant maintenance.

More specifically, wireless IoT connectivity solutions for unlicensed spectrum are expected to have significant business relevance for various industry players that typically do not own licensed spectrum. In this perspective, in 2017 the Multefire alliance is expected to start working on an IoT variant of the Multefire specifications for deployments in the ISM bands above 1GHz.

The main objective of the study is to provide a comprehensive analysis of IoT connectivity solutions for industrial applications using unlicensed spectrum. This will include (but may not be limited to):

- Study and design of contention-based uplink solutions for low-complexity IoT devices in ISM bands (with focus on 2.4 and 5 GHz bands).
- Performance characterization of end-to-end IoT performance based on available radio measurements in a realistic industrial environment.
- Potentially study the impact of inter-system interference in a controlled industrial environment.

One of the three usage scenarios identified for 5G is Massive Machine-Type-Communication (mMTC), which is characterized by a very large number of connected devices typically transmitting a relatively low volume of non-delay-sensitive data, see 3GPP Technical Report 38.913. mMTC reflects well-known use cases such as smart metering for the remote measurement and monitoring of utilities applications such as water, gas, and electricity. For those types of services, a large amount of control plane (CP) signaling may be required to tear up and down the short transmissions of low data-rate. The large CP signaling is deemed to be one the main challenges of MTC devices today rather than their user plane demand. In order to meet this challenge, 3GPP has been developing more efficient CP signaling for MTC. Recent examples of such evolution are Release-13 LTE eMTC and NB-IoT.

The objective of this study is to perform an in-depth study of CP signalling of M2M in live mobile networks covering different MTC technologies in 2G, 3G, and 4G systems in order to provide guidelines to the 5G design for efficient mMTC support.

Relevant CP signaling relates to the access connections to the network (e.g., attach/detach) and mobility and session management of an established connection. The study will include:

• Manipulate anonymised recorded traces from various datasets.
• Identify MTC devices from the traces per technology (2G, 3G, and LTE) and analyse the amount of CP signaling generated by those devices.
• Identify potential CP signalling related misbehaviour, e.g. MTC devices which cause a significant larger signalling load per technology.
• Provide recommendations on how to solve the identified issues in 5G/New Radio design.

Project:

In 2010 the theoretical capacity limit of a single-mode fiber was found. The forecasted demand of data traffic will exhaust the capacity of the installed fiber plant by 2020. New fiber-optic designs are required in order to guarantee an efficient and sustainable increase in terms of cost and energy consumption for the following 30 years. Spatial division multiplexing in its two flavors, multi-mode and multi-core fibers, has become the new research paradigm. The scope of the project is to investigate on advanced transmission and detection techniques to exploit the capacity of the new optical fiber.

• Bell Labs – Future Impossible (Youtube Channel) – The Shannon Limit
• Winzer, P.J., "Spatial Multiplexing in Fiber Optics: The 10X Scaling of Metro/Core Capacities," Bell Labs Technical Journal , vol.19, no., pp.22,30, 2014.
• René-Jean Essiambre, Gerhard Kramer, Peter J. Winzer, Gerard J. Foschini, Bernhard Goebel, "Capacity Limits of Optical Fiber Networks", IEEE-OSA Journal of Lightwave Technology, vol. 28, no. 4, pp. 662-701, February 2010.

Company: Bell Laboratories (Alcatel-Lucent) www.bell-labs.com

Bell Labs has helped weave the technological fabric of modern society. Since its founding in 1925, technology from Bell Labs has shaped the ways people live, work and play. Over the past 80 years, the Bell Labs R&D community has made seminal scientific discoveries, created powerful new technologies, and built the world's most advanced and reliable networks:

• The Transistor (1947)
• Shannon’s Information Theory (1948)
• Laser (1958)
• Communications Satellites (1962)
• The CCD (1969)
• Unix Operating System and C Language (1969-1972)
• Digital Signal Processor (DSP) (1979)
• Optical WDM systems and networks (1990)
• First 100G Ethernet transmission (2005)
• Fiber-optic capacity limit (2010)

Excellent academic records, good English knowledge, fiber-optic communications, digital communications, signal processing, VHDL programming, Matlab programming.

791 Holmdel-Keyport Rd., Holmdel, NJ 07733, USA

maps.google.com: 40°23'25.8"N - 74°11'12.7"W

70 Km from Manhattan (New York City) www.nycgo.com

Public Transport www.njtransit.com

Newark Int. Airport www.panynj.gov/airports/newark-liberty.html

• Willing to learn in a fast-paced environment.
• Knowledge in 2 or more of the subjects listed above.
• Comfortable with programming.

• Investigate existing and upcoming protocols
• Break down power consumption vs. distance / speed / mesh-size / ....
• Fine related best practices (e.g. communication to electronic price tags, room thermostats, …)
• Generate a "tool" (spreadsheet, Online tool, ...) to optimize mesh-size, gateway-density, latency, ....
• Nice to have: Investigate Fall-Back scenarios (Switch to alternative configuration immediately to ensure QoS, then optimize during operation)
• Nice to have / depending of Hardware-Affinity of Student: make a demo