Challenge 2 in FP7 ICT Call 9 Information Day Tel Aviv, 31 January 2012 Libor Král, Head of Unit Unit E5 - Cognitive Systems, Interaction, Robotics DG Information Society and Media European Commission http://www.cognitivesystems.eu
Outline Cognitive systems and robotics in FP7 Current project portfolio ICT Call 9 Tips for proposers 2
Cognitive Systems and Robotics in FP7 (2007-2012) 2012) Work Programme 2007-2008 2009-2010 2011-2012 Objective Call (Evaluation) Budget ICT-2007.2.1 (ICT- 2007.2.2): Cognitive Systems, Interaction, Robotics ICT-2009.2.1: Cognitive Systems and Robotics *) FOF.ICT.2010.1: Smart Factories: ICT for Agile and Environmentally Friendly Manufacturing ICT-2011.2.1: Cognitive Systems and Robotics Projects: ACS & Robotics (total) ICT Call 1 (2007) 96 M 17 (27) ICT Call 3 (2008) 97 M 17 (23) ICT Call 4 (2009) 73 M 19 ICT Call 6 (2010) 80 M 22 NMP-ICT-FoF (2010) 35 M 3 (8) ICT Call 7 (2011) 73 M 16 ICT Call 9 (2012) 82 M?? *) No more interaction since 2009 - language-based interaction in a separate objective with its own budget 3
Cognitive systems and robotics in FP7 Current project portfolio ICT Call 9 Tips for proposers 4
Current project portfolio ROBUSTNESS AUTONOMY ADAPTIVITY REAL-WORLD UNDERSTANDING UNDERSTANDING RECOGNISING INTERPRETING ADAPTING PLANNING MODELLING COGNITIVE ARCHITECTURES PERCEIVING TOUCHING SEEING HEARING DISTRIBUTED SENSING ADVANCED SENSING PERCEIVING APPLICATION AREAS AERIAL UNDERWATER INDUSTRY and MANUFACTURING COGNITIVE SYSTEMS & ROBOTICS LEARNING PROFESSIONAL AND DOMESTIC SERVICES MEDICAL AND REHABILITATION MONITORING AND SURVEILLANCE ACTING ACTING MANIPULATING NAVIGATING INTERACTING COLLABORATING MONITORING http://cordis.europa.eu/fp7/ict/cognition/projects/areas-projects_en.html projects_en.html 5
Call 6 + FoF mapped to SRA FOF CustomPacker/Tapas/Robofoot flexible object manipulation, packaging industry, medical applications (eg prosthetic devices) (TOMSY) Robot-arm/ML (ComPLACS) Assembly (DARWIN/Intellact) SENSOR NETWORK/MONITORING/CONTROL INTELLIGENT BUILDING (isense) SURGICAL ROBOT (ACTIVE I-SUR) LOGISTICS (RobLog) ROBOT BUILDER (Goal-Leaders) UNDERWATER (NOPTILUS / CoCoRo) Search& Rescue/ Unmanned (EMICAB) UAV ML (ComPLACs) EFAA: Game/table-top CORBYS: MOBILE MANIPULATION / esmcs / CORBYS: JAMES: social HRI - Bar tender REHAB VR of Space TOMSY: computer animation Xperience: HOUSEHOLD HELPER manipulation(intellact) esmcs: Gaming for Rehabilitation NeuralDynamics: HR COOPERATION TABLETOP SpaceBook: Navigation support robotics arm RUBICON: assisted living /transport of goods V-Charge: autonomous valet parking ecar 6
Call 7 outcome SRA mapping Manipulation of deformable objects - folding/sorting clothes (CLOPEMA) Mobile manipulation in a restaurant (RACE) Cognitive + Flexible robots for SMEs manufacturing (SMErobotics) Assistive robotics surgery (uralp, STIFF-FLOP) Underwater - cooperative/advanced autonomy (MORPH, PANDORA) Aerial cooperative manipulation (ARCAS) Outdoor guide navigation in crowded environment + social HRI (FROG) Intelligent prosthesis (CYBERLEGS) Object hand-over (CogLaboration) Compliance & Modular design (MYOROBOTICS) Safe Physical HRI (SAPHARI) Learning domestic tasks from web (ROBOHOW.COG) Understanding and carrying out everyday activity generalisation/learn action from spoken instruction/languageaction link (POETICON++) 7
Cognitive systems and robotics in FP7 Current project portfolio ICT Call 9 Tips for proposers 8
Challenge 2 in FP7 - ICT Call 9 OBJECTIVE: PUBLICATION: 18/01/2012 DEADLINE: 17/04/2012 INDICATIVE BUDGET: 2.1 Cognitive Systems and Robotics 82 M M IP: 40-56 M, M, STREP: 24-30 M, M, CA 2 M M Target (b): Cognition and control in complex systems Target (c): Gearing up and accelerating crossfertilisation between academic and industrial robotics research Target (e): Speeding up progress towards smarter robots through targeted competitions 9
Challenge 2 in FP7 - ICT Call 9 Goals Improve competitiveness of European industry Better serve the community (research community & European industry & European citizens) How to demonstrate that the field is progressing? make sciences meeting the industry and industry helping scientific progress? create a snowball effect between the offer from sciences and the needs from industry? transfer scientific results to real life? improve take up of research results (e.g. in service robotics)? 10
Target (b) IP/STREP: Cognition and control in complex systems Enabling technologies based on the acquisition and application of cognitive capabilities (e.g., establishing patterns in sensor data, classification, conceptualisation, reasoning, planning) Enhancing the performance and manageability of complex multi-component and multi-degree-offreedom artificial systems, also building on synergies between cognitive systems and systems control engineering. 11
Target (b): Expected impact Integrated and consolidated scientific foundations for engineering cognitive systems under a variety of physical instantiations. Significant increase in the quality of service of such systems and of their sustainability in terms of, for instance, energy consumption, usability and serviceability, through the integration of cognitive capabilities. Innovation capacity in a wide range of application domains through the integration of cognitive capabilities. 12
Target (b): More specifically Cognition refers to implementing cognitive capabilities in complex systems. establishing and recognising patterns in sensor-generated data prerequisite to higher-level operations such as scene interpretation, reasoning, planning, intelligent control, and complex goal-oriented behaviour learning, in appropriate modes, is essential at all levels Control refers to the general aim of making a system behave or perform in a desired way Combining cognition and control should make a system more resilient and more autonomous in the presence of real-world uncertainties and surprises. No specific approach is either favoured or excluded a priori. 13
Target (b): STREP or IP? STREPs: high-risk endeavours, breaking new grounds, with high potential rewards; also appropriate for component-level research for particular domains. IPs: for system-oriented efforts; they are expected to encompass all stages of the research and development lifecycle and, where appropriate, cutting across research topics 14
Target (c) - IP: Gearing up and accelerating cross-fertilisation between academic and industrial robotics research synergies between respective research agendas through joint industrially-relevant scenarios, shared research infrastructures, joint small- to medium-scale experimentation with industrial platforms and implementation of comparative performance evaluation methodologies and tools. 15
Target (c): Expected impact In particular: Improved competitive position of the robotics industry in existing and emerging markets for instance in the following sectors: manufacturing; professional and domestic services; assistance and co-working, production, logistics and transport, construction, maintenance and repair, search and rescue, exploration and inspection, systems monitoring and control, consumer robotics, education and entertainment.*) Also: strengthened links between industry and academia; more widely accepted benchmarks; consensus by industry on the need (or not) for particular standards. *) Note: this list is not exhaustive. 16
Target (c): More specifically It aims specifically at encouraging industry (mainly robotic manufacturers, but also users) and research institutions to work together more closely on an operational level, developing small research projects and to support technology transfer with tangible results. The projects could allow for joint enabling technology development, technology transfer, feasibility demonstration and pilots. Experiments can be proposed by industry, academia, or both. Technology transfer, mutual exchange and benefit for all participants are key ingredients to be clearly spelled out in the proposal. 17
Target (e) - CA: Speeding up progress towards smarter robots through targeted competitions based on suitably evolving reference scenarios focused on capabilities involving relevant stakeholders events, dissemination and public awareness measures 18
Target (e): Expected impact Greater innovation Measure and compare progress Increased visibility 19
Target (e): More specifically Primarily target researchers (not an engineering exercise) Industrial dimension is a plus Private sponsorship is welcome Guarantee open participation and transparent governance, demonstrate the openness and fairness of the process Spell out clearly evaluation mechanisms and performance criteria, clarify on which basis the competition will select the benchmarking Prize money not excluded, but appropriate to the circumstances (EU funding should not be used as direct cash prize) Extension of existing competitions is acceptable but there must be clear added value & EU dimension (justify the need for EU funding) 20
Cognitive systems and robotics in FP7 Current project portfolio ICT Call 9 Tips for proposers 21
Tips for proposers impact Check the expected impacts in the WP text. Any individual proposal is not expected to address the whole list, but they should clearly stress how they aim at achieving scientific/technical impacts and/or socio-economic impacts. Be concrete about which project results are likely to lead to innovation and impact and which concrete actions will be carried out during the project to achieve such impact. 22
Tips for proposers methodology Explain WHY the research is needed, WHAT will it do exactly and HOW you can progress beyond the state of the art. Not explaining sufficiently HOW a project will reach its goal is the No 1 reason for proposal failure. Avoid including an extensive literature survey within the proposal, unless to investigate a completely new research topic or sub-topic. High level requirements and specifications should be defined in the proposal. Define milestones as major project-level achievements and include measurable criteria which allow to verify the success of the RTD work which contributed to them. Handling of potential ethical issues is critical. Check the rules at http://cordis.europa.eu/fp7/ethics_en.html. 23
Tips for proposers demonstrations and benchmarks Include demonstrations/test scenarios To illustrate in sufficient details what can be realistically achieved during the project Useful to measure scientific progress Simulations are not encouraged, except for the preparation of real-world demos. Provide metrics and benchmarks Evaluation and benchmarking metrics to demonstrate progress and compare results with state of the art approaches. Define reliable criteria for assessing progress - make them public and, if possible, compare with existing criteria. Show how to measure performance - objective evaluation of key system properties. 24
Tips for proposers integration Integration is a major challenge, in particular for projects with different disciplines, different types of hardware, cooperating systems, novel combinations of components etc. Especially if your proposal addresses more downstream robotic systems prototypes ensure you describe concretely how you will achieve systems integration. TIP: Lack of attention to project integration is a major cause for failure. 25
Tips for proposers follow ups and resubmissions Follow up of successful previous projects: Above all, explain clearly the new added-value of the follow-up. It should not be a mere continuation. Partial or full resubmissions of previously rejected proposals are allowed, and treated like all other (new) proposals. But you are strongly advised to take into account the context and aims of the new Call for Proposals and if still relevant, the comments from the previous ESR where applicable to your new proposal. 26
Tips for proposers budget considerations Demonstrate project s value for money by spelling out the costs of and need for personnel, equipment and other resources as being commensurate with the project scope and aims. Choice between a STREP or an IP in target (b) should be based on the scope and content of the project rather than the size or budget. No upper limits are applied and IPs do not need to have a large consortium. 27
Tips for proposers project partners Select consortium partners for the job in hand, based on their qualifications, scientific track record (include their top few publications), and their ability to complement each others roles, skills and experience, compared with the expertise needed to achieve the project's goals. Make sure to have enough partners to do the job but not more than needed. Geographical balance is definitely NOT an evaluation criterion. Select your partners based on the needs of the project, not on their location. The minimum requirement is three mutually independent partners from three different countries (check the FP7 Rules of participation for the exact terms). 28
Tips for proposers industries and end users Industrial partners can have several potential roles carry out research provide infrastructure/tools provide challenging application scenarios as source of research questions give access to their facilities or real test-cases scenarios in order to assess scientific progress. In all the cases, they have to show they are stakeholders in the work and its results and demonstrate commitment and strategy to exploit the project results. End users are often not required in scientific research projects but if they provide clear added value, end user participation, particularly in more downstream systems integration projects, is very welcome. 29
Tips for proposers dissemination Projects with a longer-term, scientific or academic bias are expected to have a very pro-active and well targeted dissemination plan and include a high number of public deliverables and public-domain results. Projects with more industrial dimensions do not necessarily need to make all deliverables public. The proposal should make this clear and describe its plans to protect or exploit such results e.g. through patents, licensing or technology transfer. In both cases, ensure you describe the main target audiences, venues and channels for dissemination / exploitation. 30
Tips for proposers exploitation Projects with a more industrial approach should concretely describe the path towards exploitation (including joint exploitation by the partners) of transferable results from the work. However, developing business plans is beyond the scope of this RTD funding programme. Projects with a more scientific approach should still make clear what the eventual exploitation outcomes and impact on the scientific community will be, including re-use of scientific results in further research. In either case, exploitation strategy should include identifying potential exploitable results and target users, as well as a mechanism for attracting them. 31
Further information Call info including pre-proposal feedback service: http://cordis.europa.eu/fp7/ict/cognition/calls-ictcall9_en.html Q&A document: http://cordis.europa.eu/fp7/ict/cognition/docs/call9faq.pdf Information about existing project portfolio: http://cordis.europa.eu/fp7/ict/cognition/projects/areasprojects_en.html News: http://twitter.com/roboticseu http://www.facebook.com/roboticseu 32
THANK YOU!