Seminarios 2018-2019

Título: Artificial Intelligence and Games

Ponente: Diego Perez-Liebana

Fechas: 13 y 14 de diciembre de 2018

Horario: de 10:00 a 13:00h

Organizador: Pedro Isasi

Lugar: Salón de Grados. Campus de Leganés

Créditos: 1 ECTS

Resumen (indicando contenido por sesiones): 

Games have always been excellent benchmarks for the advancement of AI. From Machine Learning to Search and Decision-Making algorithms, from Deep Blue and Alpha Zero to Starcraft 2, researchers across the globe agree on frameworks to test their methods. This course gives an overview of the Game AI field, focusing on the different research opportunities that it offers and some of its applications.

This seminar will cover the different ways AI research is performed in this community, in terms of creation of intelligent agents that play games but also using AI to create content for games and influence game design. Particular focus will be put on the i) algorithms that have revolutionised the field in the last decade, such as Monte Carlo Tree Search and Deep Learning; ii) the most known frameworks that are being used nowadays; and iii) the most promising lines of future work for the next decade in this field.

Breve biografía:

Diego Perez-Liebana is a Lecturer in Computer Games and AI at QMUL and holds a PhD in CS from the University of Essex (2015). He has a BSc and MSc from University Carlos III (Madrid, Spain). His research interests are Search Algorithms, Evolutionary Computation and Reinforcement Learning applied to Real-time Games and General Video Game Playing.

He's published more than 60 papers in leading conferences and journals in the area. He's the main organizer behind popular AI game-based competitions in the field and the general chair of the upcoming IEEE Conference on Games (QMUL, 2019). He has experience in the videogames industry with titles published for both PC and consoles, and also developing AI tools for games. He has also secured private funding from the industry (Microsoft Research and Google Deepmind) for the development of game AI competitions.

Título: Engineering challenges and opportunities in dealing with the Cyber-Physical Systems of Systems of tomorrow

Ponente: Martin Törngren, Ph.D, Department of Machine Elements, KTH

Fechas: 21, 22 y 23 de Enero de 2019

Horas: de 15:00 a 19:00h

Organizador: Anabel Fraga Vázquez

Lugar: 3.S1.08 Edificio Biblioteca Rey Pastor. Campus de Leganés

Créditos: 2 ECTS

Resumen:

Cyber-Physical Systems (CPS), as integrations of computation, networking and physical processes,  are becoming autonomous, smart, connected and collaborating, resulting in the creation of Cyber-Physical Systems of Systems (CPSoS) with unprecedented capabilities and opportunities.

The implications are that CPS will be widespread and that a multitude of key societal functions (such as water, energy, transportation, and health-care) will be relying on the proper operation of such CPS. In a series of lectures I will elaborate on

• (i) What we mean by CPS:
  - providing etymology, definitions, examples, characteristics, and discussing trends
  - relating CPS to other terms such as IoT, IIoT and I4.0
   
• (ii) CPS opportunities and challenges, with examples from several industrial domains,
  - generic trends across multiple application domains
  - sociotechnical considerations and implications for CPS
  - elaborating complexity facets of CPS, including characteristic differences among software and physical parts within CPS, 
  - discussing limitations of existing methodologies, of humans, organizations and computer-aided engineering systems used to develop CPS. 
  - the role of systems engineering for future CPS
   
• (iii) Industrial revolutions and the role of CPS in this context:
  - what characterizes and industrial revolution - and are we embarking on one?
  - push and pull factors
  - What are the implications - which domains will be affected, how, and why? 
  - For which domains will market impact be radical or evolutionary? 

While several industrial examples will be given, for illustration purposes there will be an emphasis on automated driving and intelligent transportation systems.
An active style of lecturing will be adopted, with work group sessions integrated into each part.

Related reading material:

• CyPhERS CPS agenda: http://cyphers.eu/sites/default/files/d6.1+2-report.pdf
• Complexity Challenges in Development of CPS, Book chapter Springer: Principles of Modeling 
• Essays Dedicated to Edward A. Lee on the Occasion of His 60th Birthday 
  https://link.springer.com/chapter/10.1007/978-3-319-95246-8_27

Breve biografía:

Martin Törngren has been a Professor in Embedded Control Systems at the Mechatronics division of the KTH Department of Machine Design since 2002. He has particular interest in Cyber-Physical Systems, model based engineering, architectural design, systems integration, and co-design of control applications and embedded systems. He has authored/co-authored more than 100 peer reviewed publications, and also been in charge of developing and leading graduate and continued education courses. He spent time as a post-doc at the EU-JRC, and did a 10 month sabbatical 2011/12 at UC Berkeley. In 1996 he co-founded the company Fengco Real-time Control AB, specializing in advanced tools for developers of embedded control systems and related consultancy. In 1994 he received the SAAB-Scania award for qualified contributions in distributed control systems, and in 2004 the ITEA achievement award 2004 for contributions in the EAST-EEA project. He served as the technical coordinator of the international iFEST ARTEMIS project with 21 partners (2010-2013).  

Networking and multidisciplinary research have been characteristic throughout his career. From 1999-2004 he served as the Chairman of the Swedish real-time systems association, and he has represented KTH as a core partner in the EU networks of excellence in Embedded systems design, Artist2 and ArtistDesign, and in the Artemis industrial association. He is moreover the principal initiator and Director of the Innovative Centre for Embedded Systems (www.ices.kth.se), launched in 2008. 

Personal web page: https://www.kth.se/profile/martint

Título: Performance programming on Android devices

Ponente: Prof. Rafael Asenjo, Universidad de Málaga

Fechas:

• Sesión 1: 11 de Febrero de 2019
• Sesión 2: 12 de Febrero de 2019
• Sesión 3: 13 de Febrero de 2019
• Sesión 4: 14 de Febrero de 2019

Horas: de 10:00 a 13:00h

Organizador: Francisco Javier García Blas y José Daniel García, Grupo ARCOS, Departamento de Informática

Lugar:

• Lunes, 11 de Febrero:  2.2.C06
• Martes, 12 de Febrero:  2.2.C06
• Miércoles, 13 de Febrero:  2.2.C06
• Jueves, 14 de Febrero:  4.0.F16

Créditos: 2 ECTS

Resumen:

It is certainly challenging to stand out in the buoyant market of mobile applications. In many cases, the mere fact that your application shows a more fluid execution and/or consumes less battery can be key for a better positioning in the rankings of applications.

In this talk we cover different aspects of programming for Android devices that will help you improve the performance of your apps. More specifically, we will comment on the increase in performance that you can achieve using the NDK (Native Development Kit) and exploiting the architecture of ARM processors (multicores, intrinsic NEON, ASM and GPUs). This course covers from the study of the architecture in the processors that drive modern mobile devices (smartphones, tablets, etc) to the study of code optimization techniques for applications that run on such devices.

* Introduction. ARM processors (1h)

* Android programming

• Android main concepts (2h)
• Layout, listeners and life cycle (3h)
• Camera UI and Image Processing (1h)
• Native Code (NDK and JNI) (2h)

* Optimizations with Parallel libraries

• Parallel Java basics (3h)
• Threads C++11 and OpenMP on Android(3h) Learning outcomes:

The student will be able to:

1. 1. compare and assess different architectures available in the market and targeted to mobile devices.
2. 2. evaluate the Hardware/Software trade-offs when optimizing the programming of a mobile device.
3. 3. handle programming environments and frameworks for the development of applications for mobile devices, as well as  optimize these applications

Breve biografía:

Rafael Asenjo obtained a PhD in Telecommunication Engineering from the U. of Malaga, Spain in 1997. From 2001 to 2017, he was an Associate Professor in the Computer Architecture Department, being a Full Professor since 2017. He collaborated on the IBM XL-UPC compiler and on the Cray's Chapel runtime. In the last five years, he has focused on productively exploiting heterogeneous chips.

He has had the opportunity to teach degree level subjects related to Technology, Structure and Architecture of Computers, Operating Systems, Real Time and Emerging Architectures. He also teaches in the Master in Mechatronic Engineering (Multiprocessor Programming), in the Master in Telecommunication Engineering (Lab. of Architectures for Mobile Devices) and in the Master in Computer Engineering (Architecture and Programming of Mobile Devices). In the last two subjects mentioned, the optimization of the performance of Android applications is deepened. Rafael Asenjo is also a qualified instructor for the Cisco CCNA certification.

Título: Recent advances in automated planning

Ponente: Amanda Coles

Fechas: 25 y 26 de octubre de 2018

Horario: de 10:00 a 13:00h

Organizador: Susana Fernández

Lugar: 3.1.S08 Edificio Biblioteca. Campus de Leganés

Créditos: 1 ECTS

Resumen (indicando contenido por sesiones): 

En este seminario se tratarán temas relacionados con Planificación automática de tareas, especialmente relacionada con su variante temporal.

Breve biografía:

PhD in Artificial Intelligence Planning, Department of Computer and Information Sciences, University of Strathclyde. Start date: October 2003, thesis submission September 2006. Conferred July 2007. My thesis, entitled "On The Inference and use of Macro-Actions in Forward Chaining Planning", can be found on my publications page BSc (Hons) Artificial Intelligence, First Class, University of Durham, October 2000 to July 2003.

I am a lecturer in the Department of Informatics at King's College London, and am part of the Planning Group at King's. My research interests lie in the area of AI planning, I am particularly interested in hybrid systems planning (planning with discrete and continuous numeric change); the optimisation of plans to satisfy preferences (soft constraints) and generating plans of high utility in situations with limited resources. I have a strong interest in the application of AI planning to real world problems, and as such have also developed techniques for reasoning with highly expressive world models, including reasoning about continuous numeric resources. I am currently principal investigator on the EPSRC funded project "AI Planning with Continuous Non-Linear Change" (EP/P008410/1) and co-investigator on the European Comission project ERGO (European Robotic Goal-Oriented autonomous controller).

From April 2010 to April 2013 I held an EPSRC Post Doctoral Research Fellowship (EP/H029001/1) to research planning in problems with continuous numeric change, where there is uncertainty about the resource consumption and duration of actions. My work on planning with continuous numeric resources, began as part and earlier EPSRC project (EP/D062721/1) at the University of Strathclyde.

Recent major contributions of my work include development of the planners: Colin, capable of reasoning with continuous numeric change; LPRPG, discrete numeric change using an LP to perform complex numerical reasoning; and LPRPG-P an extension of LPRPG making use of the LP as an optimisation tool to create plans satisfying preferences. The planner POPF, which extends Colin to allow partial commitment in forward search temporal planning, is the most recent success having been awarded the runner up prize in the temporal track of the 2011 International Planning Competition. POPF was the only planner in this competition to support the full range of temporally expressive planning problems.

Temporal and numeric reasoning is vital in allowing planning to be applied to real world problems, which contain much of this type of structure: charge in Martian Rovers, Fuel in Logistics and Power Demand in Electricity supply. My research interests therefore extend to application of planning to real-world tasks, and through this I have been involved in developing VOLTS in conjunction with the Department for Electronic and Electrical Engineering at the University of Strathclyde. VOLTS is a system for asset management based on the Grendon Substation, part of the UK's National Grid, near London. I have also been involved in work with Rune Moller Jensen and Kevin Tierney of ITU Copenhagen on automated planning of ocean-liner repositioning for the shipping company MAERSK.