FRANCE STATUS REPORT
I - STATUS OVERVIEW
1.1. INTRODUCTION
Following the Williamsburg summit held in June 1983, IARP encompasses research activities on Advanced Robotics with the principal aim to develop robotics technology for hostile or difficult work environments. This covers indeed very vast domains that can be characterized at two levels.
First, at the level of basic research, it fosters a host of technologies ranging from the design of very novel mechanical devices such as : locomotion architectures, microsystems, dextrous hands,... to the most advanced techniques in machine intelligence such as : 3D environment perception and modelling, motion and task planning, reactivity and decisional autonomy,...
Second, a broad host of new real-world applications comprising :
- humankind frontier fields ranging from deep-sea work to outer space,
- socially highly relevant subjects such as : health care, personal assistance robots for the impaired, mining, agriculture,...
In this status report, we will cover research efforts in France associated with this objective both through national actions and via cooperative international programmes among which ESPRIT and EUREKA play a key role.
The main actors in France are national research organizations. Among them, CNRS (Centre National de la Recherche Scientifique) and INRIA (Institut National de Recherche en Informatique et Automatique) in close relation with university laboratories, pursue the broadest spectrum of in-depth studies in basic research such as : manipulation, navigation and motion planning, perception and multi-sensor fusion, geometric modelling, real-time knowledge-based systems, learning, task planning and temporal reasoning, control architectures, etc.
Several domain-oriented organizations carry on important R&D efforts which bear direct impact on the overall IARP domain and are such that their activities are part of this report. They are, by alphabetical order : CEA (Commissariat à l'Energie Atomique), CEMAGREF (Centre National du Machinisme Agricole du Génie Rural, des Eaux et des Forêts), EDF (Electricité de France), CNES (Centre National d'Etudes Spatiales), IFREMER (Institut Français de Recherche pour l'Exploration de la Mer), ONERA (Office National d'Etude et Recherche Aérospatiales).
1.2. CURRENT ACTIVITY OVERVIEW
The previous mentioned organizations cooperate in many ways to foster Advanced Robotics and to enhance the synergy and increased efficiency of the French research teams engaged in the domain. This is a major characteristic, clearly apparent in most of the projects and programmes hereby reported.
In particular, this has lead CNRS to organize a cooperative programme, the Strategic Action on Intelligent Machines, that creates a national network of domain-oriented projects. All those projects are structured to include CNRS laboratories in partnership with appropriate public institutions as well as industrial partners (cf. section 2.1.1.).
Similar objectives and efforts are carried on by INRIA within a set of nine research projects (cf. section 2.1.2.)
We will report on the status of these important endeavours within the part of the report devoted to National Programmes (cf. section 2.1) that will include an up-dated presentation of a major joint R&D action The Engineering Science Consortium Initiative. Section 2.2 will also briefly report on other current on-going national and international programmes and projects. We will consider domain by domain the most relevant research work and projects within the IARP field. The scope is limited to already established or emerging cooperative work with special emphasis on international cooperation.
In the specific sections 2.3. to 2.8., we will report on the activity in the following application oriented domains :
• Health care and Human Assistance Robots • Agriculture and the Food Industry
• Nuclear Robotics
• Human Assistance Robotic Systems
• Subsea robots
• Field robotics
2. R&D ACTIVITY REPORT
2.1. NATIONAL PROGRAMMES
2.1.1. CNRS Strategic Action on Intelligent Machines
2.1.1.1. Action overview
The assessment of the current development of Advanced Robotics, supported by information technology, microtechnologies and microsystems developments, has lead to consider for Robotics the new frontier of non-manufacturing applications. Indeed, sensor based systems and embedded intelligence are decisive factors for a host of new machinery and products that will take the robot out of the floorshop :
- to the service-robots domain
- in field-based applications such as mining, forestry, agriculture, underwater, space,...
- to public-oriented areas ranging from domestic and professional cleaning to assistance to the disabled and the aging
Indeed, this new frontier for Advanced Robotics offers a double challenge and a double opportunity.
First, the challenge and the opportunity to offer solid grounds for Robotics to emerge as an application-independent field, to construct it as a true scientific discipline based on a coherent set of concepts and formal tools. This implies the actual development of machine-intelligence based robots, i.e. systems that would have to integrate perception to action throughout levels of sensing interpretation, task assessment, symbolic reasoning, sensor-based execution control, in a fully integrated system that will be controlled by a human throughout a sophisticated operator-machine interface.
Second, the challenge and the opportunity to develop application-oriented robots and systems in a broad host of domains with attributes that will range from exciting technical achievements to meaningful economical impacts and high-social relevance.
At the national level, several laboratories are currently involved in advanced robotics projects with partners from general and domain-oriented public institutions and industry.
Accordingly, CNRS has organized a network of ten specific actions associating several laboratories to a focal point laboratory and open to cooperation with other institutions as well as to industrial partners.
2.1.1.2. Autonomous Intervention Robots for non-Cooperative Sites
Project objectives : the research programme is concerned with the design and the system development of autonomous mobile robots acting in natural and non (or partly cooperative environments with remote teleoperation and task-level teleprogramming.
The research themes include :
. Locomotion system design and control
. 3D perception and environment modelling
. Decisional real-time on-board autonomy
. Teleprogramming and telesupervision systems
The applications and on-going projects concern planetary rovers, in particular for long traverses and complex missions such as IARES (cf. section 2.8.4.), exploration and assistance robots for scientific tasks (e.g., search for meteorites) or the Antarctic base CONCORDIA (cf. section 2.8.5.), demining robots, etc. Several topics are investigated One main issue is navigation in unknown natural environments for which an approach adapting the complexity of the models and computing processes to the complexity of the terrain (flat/uneven/cluttered) is developed to comply with energy and computing power limitations. Another issue is the coordinated exploration or task achievement by several robots. This includes environment model sharing, coordinated planning and execution, and mutual localization, and is particularly of importance for the demining and exploration applications.
CNRS focal laboratory: LAAS
Other CNRS laboratories: LIFIA, LIRMM, LRP
Testbeds: LAAS facilities, GEROMS (cf. section 2.8.4.)
Main partners: CNES, RISP (cf. section 2.8.7.)
Industrial companies: Alcatel Espace, Cybernetix, ITMI, Matra Marconi Space, SAGEM
2.1.1.3. Construction and Civil Work Robots
Project objectives: the research programme comprises the system developments implied for a mobile machine to operate in construction works (road construction and maintenance, building construction,...).
It includes subjects such as : 6D dynamic localization, navigation and operator machine interfaces for soil automated handling, obstacle detection in complex dynamic work-environments.
CNRS focal laboratory: IRCyN
Testbed : LCPC site: Robot Melody, ...
Institutional partners: LCPC, CER (roadway experimentation center), CEMAGREF.
Laboratories: ISIR, LASMEA, CAOR,LRP
Industrial companies: Cochery Bourdin Chaussée CBC, Scetauroute,ITMI,
Europeen Project CIRC: BRITE-EURAM N° 96-3039
Duration: three years (1997-1999)
CIRC Partners: ITMI APTOR (coordinator), LCPC, Universty of Karlsruhe (Germany), CBC,TEKLA OY (Finland), National Land Survey (Sweden), University of East London (UK).
2.1.1.4. Machine-Intelligence Assisted Driving
MAN-MACHINE COOPERATION FOR DRIVING ASSISTANCE
Focal Laboratory: HEUDIASYC
Participants
Laboratories
CAL (Lille)
HEUDIASYC (Compiègne)
IEF (Orsay)
LAAM (Paris)
LAAS (Toulouse)
LAMIH (Valenciennes)
LASMEA (Clermont-Ferrand)
LM2S (Troyes)
National Institutes
INRETS
DART-CIR ( Department of traffic Analys and Regulation)
LPC (Lab of Psychology of Driving)
Industrials Partners
PSA
Renault
Thomson-CSF
Following the EUREKA Prometheus program, and particularly the ProArt project, laboratories with their industrials partners have defined projects centered on safety and comfort. These researches concern for a part, the determination of accident mechanisms and their models as well as studies realized in real situations allowing to analyze driver-vehicle behaviors, driver vigilance according to driving situations, driving strategies and accidents avoidance. Another aspect of the research programme deals with the development of new perception systems adapted to automobile domain as well as the design of new methodologies for developing real-time embedded applications and implementing them on heterogeneous hardware architectures.
Current projects on these different subjects are the following:
Modeling the driver behaviors, Diagnosis of Dynamic behavior of vehicles, Modeling a convoy of vehicles, Alarms transmission between road actors, Detection-location-recognition of obstacles by CCD camera, Reconfigurable boardscreen, Real-time embedded architectures.
2.1.1.5. Personal Robots to Assist the Impaired and the Aging
Project objectives : the technical issues and scientific orientations of the research programme cover the following aspects :
A first specific project illustrates the scope, the constraints and the objectives of the programme : to endow a wheelchair for apractic children who are mentally and physically disabled with advanced robotics functionalities in order to move in a structured and specially designed environment and, hence, to allow for a higher level of autonomous living.
CNRS focal laboratory : LAAS
Other laboratories : IRIT, LAEI
Testbed : mobile robots, arm equipped wheelchairs (ENIT, LAAS)
Institutional partners : INSERM (CREA-Paris ; U103, Montpellier), ENIT, CEA
2.1.1.6. Robot Assisted Surgery
Project objectives:
The research programme covers the field of computer-assisted medical intervention related to surgery. Systems for computer aided surgery must be safe both in their software and in their hardware components. The project approach based on the use of multi-modality medical data, concerns passive and semi-active systems guiding visually or mechanically the surgeon towards/within a planned surgical strategy. Major interest of the partners concerns:
CNRS focal laboratory: TIMC-Institut de Mathématiques Appliquées de Grenoble and Grenoble University Hospital
Other laboratories: LASMEA (Clermont-Ferrand), INSERM LTSI (Rennes), LATIM (Brest), GRAVIR (Grenoble), CLIPS-IMAG (Grenoble), LIME (Grenoble), LIFL (Lille), EPIDAURE-INRIA (Sophia-Antipolis), LIRMM (Montpellier)
Testbed: Technical experimentation room at TIMC and Surgical room of
Grenoble hospital (CHU)
Institutional partners: INSERM, INRIA
Industrial partners: SOFAMOR, ICP-AESCULAP, PRAXIM, HEXALE, CERTIS, ALLOPRO,...
2.1.1.7. Robotics in Agriculture & the Food-Industry
The project objectives, the technical and scientific orientations of the research programme are being reconsidered following the new orientation of CEMAGREF (cf. 2.4.1).
CNRS focal laboratory: LASMEA, Blaise Pascal University of Clermont-Ferrand
Other CNRS laboratories: IRISA (University of Rennes 1), LIRMM (University of Montpellier 2), HEUDIASYC (University of Technology of Compiègne)
Institutional partner : Cemagref (research groups in Clermont-Ferrand, Montpellier, Rennes and Antony).
The technical issues and scientific orientations of the research program cover the fields of underwater remote intervention and ocean exploration. This can be envisioned in two modes: ROV (Remotely Operated Vehicles), and AUV (Autonomous Underwater Vehicles).
Initially, the program mostly focused on ROV research. Some of the most important research topics included:
. development of operator machine interfaces (augmented reality,...)
. vehicle dynamic stabilization under heavy environment perturbations,
. redundant on-board arm teleoperation including force feedback control,
. vehicle teleoperation with time delay,
. sensor based control,
. arm-vehicle coordination.
The experimentation site is located at IFREMER (Institut Francais de Recherche pour l'Exploitation de la MER), where the ROV VORTEX is fitted with a 7 d.o.f. arm (Mitsubishi PA-10).
The scientific need for new oceanographic data acquisition tools, however, has also oriented underwater research towards Autonomous Underwater Vehicles. The LIRMM (Laboratoire d'Informatique, de Robotique et de Microelectronique de Montpellier) currently develops a small size and low cost AUV named TAIPAN, based on an acoustic target developed by CNIM (Constructions Industrielles de la Méditerranée) and used for
anti-submarine warfare training. This AUV is available to the scientific community as a testbed for research on: navigation, control, communications, telemanaging, data acquisition, etc...
CNRS focal laboratory: LIRMM
Test bed: Toulon IFREMER site and LIRMM Montpellier
Other CNRS laboratories: IRCyN, I3S, LAG, LAAS, LASMEA.
Main partners: IFREMER, INRIA, and RSM (Mediterranean Subsea Robotics Group) lead by CNIM.
The AUV Taipan developed by LIRMM is a small size and low cost vehicle. It is derived from the torpedo-shaped AUV, named CALAS, developed by CNIM and used as an acoustic target for anti-submarine warfare training.
It measures 1.66m in length, 15 cm in diameter, and weighs 25 kg. It is powered with sealed rechargeable lead-acid D-cells. The control electronics is based on transputer boards programmed in parallel C. A Matlab graphical interface allows to program the vehicle trajectory and to display variables of interest for analysis after completion of the mission.
Like CALAS, it has a single propeller, a rudder and a stern diving plane. An additional diving plane located in the nose, however, endows Taipan with new capabilities. These control surfaces have indeed several advantages :
. they provide extra lift for low speed control,
. they allow to dive from the surface, even in calm water,
. they endow the vehicle with the ability to change depth at zero pitch.
2.1.1.9. Teleoperation and Virtual Reality for Hazardous Sites
The project general objective is to design, experiment and validate effective Telepresence Teleoperation in the context of nuclear and similar hostile sites.
The research programme covers the subjects and issues necessary to develop enhanced teleoperation systems building on present advances in Virtual Reality technology.
Actually, classical teleoperation systems suffer from two main drawbacks: an insufficient ergonomics (difficulty for the operator to forecast the slave motion when moving the master arm or joystick) and a bad quality of sensory feedback (difficulty for the operator to understand the slave world and what happens in it). Virtual World, according to our analysis, can be a "buffer" world designed in order to be well adapted to the human operator abilities while Real slave world will be well adapted to machines or robots abilities. The main problems to be solved deal with the development of input/output tools for Virtual Reality systems (for most of them multi-sensory feedback is a main feature), with representation of tasks in Virtual Worlds (because Real Slave World and Virtual World have not the same geometry and topology), and with time delays mastering in case of long distance teleoperation.
CNRS focal laboratory: Laboratoire de Robotique de Paris (LRP).
Main institutional partner: CEA (Atomic Energy Agency).
Other laboratories: Ecole des Mines de Nantes, CEMIF (Evry University), INRIA-GRAVIR (Grenoble), LASMEA (Clermont-Ferrand.
Industrial partners: EDF (DER), THOMSON (LCR).
Testbed: A main experimental site located within the CEA (Paris) premises and connected to LRP facilities.
That programme is closely connected with several international projects: TELEBORG (EEC programme about VR aided teleoperation systems for industrial assembly ; with KUL (Belgium) and France Telecom. MAESTRO (EEC programme about VR control of robot on satellite), GALILEO (bilateral LRP - Santa Anna (Pisa, Italy) programme on haptic feedback in VR aided telepresence), TWE (CNRS PICS programme with MEL (Tsukuba, Japan) about long distance VR aided multi-teleoperated concurrent systems).
Todate, preliminary results are concerned with design and implementation of:
1) several types of 6 d.o.f. force feedback master joysticks,
2) 20 d.o.f. force feedback data glove,
3) several softwares for collision detection in virtual worlds, for analysis of force distribution in case of:
- virtual hand-virtual environment multicontact,
- progresses in weight sensory feedback when manipulating virtual objects,
- experiments in remotely controlling through a VR station four different robots located in four places in the world and performing in parallel the same assembling task,
- experimental results allowing to propose a new method to overcome the time-delay problem.
2.1.1.10. Microrobotics and Microsystems
Project objectives: the main goal of this project is to study and to develop new robots in millimetric and submillimetric dimensions. Reaching this goal involves several problems including:
- use and/or design of microactuators, microsensors,
- microrobots structure design,
- microtechnologies for the realization of such structures,
- control of the microrobotics motions in opened and closed loops,
- integration.
To study these problems under several points of view, three projects have been defined, corresponding to three different demonstrators. These projects will illustrate in an interesting way the specific scientific problems in Microrobotics.
The three projects are:
- very small pipes inspection,
- swarm microrobotics,
- flexible microassembly system.
The project "very small pipes inspection" concerns the design and realization of microrobots for the inspection of pipes with less than 15 millimeters in diameter. In this field two main problems are studied. The first one is locomotion. For the motions of the microrobot two types of microactuators are used : the shape memory alloys and the piezoelectric materials. The second problem is the inspection itself and the research concerns the development of a microcamera.
The aim of the project "swarm microrobotics" is to design a big number of microrobots (about 10) and to control them in order to perform collectivelly the transport or pushing of an object. Two aspects will be treated : the technological aspect i.e. the design of a robot of about 1 cm3-size and the collective intelligence aspect i.e. the methodologies to derive individual reflexive behavior from the defined collective behavior.
The project "flexible microassembly system" concerns the study of a microassembling station with microconveyors and micromanipulators. The whole system should be decimeter-size.
CNRS focal laboratory: Laboratoire d'Automatique de Besançon (LAB - UMR 6596)
Other laboratories: LRP (Paris), CRAN (Nancy), DCPR-Grapp (Nancy), LAAS (Toulouse), IIUN (Neuchâtel), LIMMS (Tokyo), LMA (Besançon), LMS (Poitiers), LAIL (Lyon), IEMN (Lille), TIMC (Grenoble), LSIIT-GRAVIR (Strasbourg), HEUDIASYC (Compiègne), LAFORIA (Paris), IGE (Belfort).
Main institutional partner: IMFC
Industrial partners: EDF, Thomson, General Electric Medical Systems, Framatome
Testbeds: various experimental sites in relation to three projects:
- very small pipes inspection (LAB, LMA)
- swarm microrobotics (LAB, LRP)
- flexible microassembly system (IEMN, LAIL)
2.1.1.11. The Intelligent Home
Project objectives: In a first phase, home electronics-based technology was mainly directed towards confort improvement and security enhancement. Currently, the fast growing developments of Information Technologies and Machine Intelligence have open a larger set of issues that integrate related systems within modern housing design and construction with special emphasis in multi-media and network aspects.
Hence, the new field of Intelligent Homes can encompass much more ambitious objectives. In accordance, this project has selected to concentrate the efforts in two correlated domains:
- Health telesurveillance and care, including post-illness recovery,
- Maintaining and assistance in their home environment of the aging and the handicapped.
Foreseen system developments include subjects such as smart sensors, human-machine interfaces, communication protocols, and diagnosis systems.
CNRS focal laboratory: LAAS-CNRS
Other laboratories: CRIN, ITU Blagnac,...
Institutional partners: INSERM (Toulouse), Centre Hospitalier (Nancy), Fondation Mérieux.
Industrial partner: BOUYGUES
INRIA research activities are organized in projects that interact and cooperate within the the Institute itself, and most important, with outside partners from public institutions and industry, both at the national and at the international level, in particular in the framework of European R&D activities. The following briefly describes the objective and the scientifical and technical scope of the eight projects related to the domain covered by IARP.
2.1.2.1. ICARE (Instrumentation Control and Architecture of Advances Robots)
This project is dedicated to the study of the problems of command and control in autonomous mechanical systems, with a special but not exclusive focus on robotics. The main target is to develop a methodology for the design and integration of each link in the chain.
Research themes :
… Algorithms and theoretical aspects of control.
Extending the theory through handling non-standard problems (non-holonomous system).
… Acquisition and use of sensory data. Robot autonomy capacity is linked to the utilization of exteroceptive sensors and the use of sensory data in the control loops.
… Architecture and programming for robot controllers (ORCCAD).
The complexity of the applications makes it essential to study and develop appropriate systems for programming and simulation.
2.1.2.2. BIP (Human Locomotion, Control of Biped Robots and Complex Systems)
This project began on January 1, 1994 with the aim of designing biped walking robots and their control systems. The interest in these robots lies in their natural capability for operating in the essentially bipedal-friendliness of our everyday environment. The priority class of applications will be that of service robots, with the hope of further spin-offs in the area of biomechanics.
Research themes:
… Modeling human locomotion in various configurations: measuring parameters for group of individuals, the search for postural or motion invariants, determining underlying optimization criteria.
… Study of control systems based on the active stabilization of quasi-periodic passive solutions in conjunction with redundant task and sensor-based control technologies (force, proximity and vision).
… Development of design/programming/verification tools for a reactive real-time control system (ORCCAD).
… Design of specific mechanical sub-assemblies.
… Synthesis and animation of human walking.
… Searching for limit cycles and chaotic behaviors of passive walking systems.
2.1.2.3. EPIDAURE (Medical Image Analysis and Simulation)
The goal is to design and develop new tools to analyze multidimensional and multimodal medical images (CT images, magnetic resonance images, ultrasound images, nuclear medicine images, etc.) in order to improve diagnosis and therapy, specially when therapy is guided by medical images (video-surgery, surgical radiology, radiotherapy, etc.). Also, we study user interaction with medical images, especially as part of surgical operation simulation.
Research themes:
… Extraction of quantitative parameters useful for diagnosis (shape, texture, motion), spatial registration of images acquired at different times, fusion of multimodal images, differential geometry, analysis of deformable motion, construction and use of electronic anatomical atlases, morphometrical and functional brain analysis, building virtual patients and simulation of surgery, virtual and augmented reality in medicine, spatial localization of patients and surgical tools, coupling medial imagery with medical robotics.
… Software and hardware integration.
2.1.2.4. MOVI (Geometric Modeling and Computer Vision)
The project aims to bring general and fundamental solutions in order to recognize, locate, measure shapes seen through one or several cameras, and to validate these solutions through various applications ranging from robotics to image data base consulting. The basic tools which we are developing are three-dimensional perception geometry, invariants linked to this geometry, the matching from image to image and from image to model.
MOVI is a common project with the GRAVIR/CNRS laboratory.
Research themes:
… Multiple image geometry. This point concerns the geometric constraints linking in several views.
… Camera calibration. It is a difficult technical stage which can be considered as controlled now. Current work is centered around autocalibration for a single camera or for the camera-hand calibration in robotics applications.
… Perception-action coupling. It is about the integration of perception and action aspects and the study of new problems arising when sensors are badly calibrated.
… Matching. This basic problem is explored in two ways : between images for stereovision, and between models and images, for instance for image data base retrieval.
2.1.2.5.
ROBOTVIS (Computer Vision and Robotics)
This project aims to develop the theory and practice of machine visual perception. To this end, the team is building mathematical and computational tools and testing results against real applications and performance in biological systems.
Research themes:
… In order to solve a given visual perception task, we must first answer the following questions: what information is to be extracted from images and what sort of mathematics allows this ? What are the uses of these methods and what experiments can adequately validate them ? And what computing architectures will execute these algorithms in real-time in a given application ?
… Theoretical problems: early vision, perception of distances and movements, the presentation of object shapes, the recognition of objects or places, co-operation between visual and inertial modalities, relations between learning and visual perception, active visual perception.
2.1.2.6. SHARP (Automatic Programming and Decision-Making Systems in Robotics)
The goal of this project is to develop the concepts of task-level programming and autonomy in robotics. The application of these concepts should facilitate the use of robots for specific, non-repetitive tasks, in particular, those which preclude the use of a human agent. The emphasis of the discussion lies on the areas of planning and execution of complex movements (including those related to dexterity) and reactivity.
SHARP is a common project with the GRAVIR laboratory of IMAG (CNRS, INPG, UJF).
Research themes:
2.1.2.7. TEMIS (Advanced Image Sequence Processing)
The main research dimension to this project is accounting for multi-image sequences (usually temporal sequences), as the input to the processes of analysis, compression or interpretation ; motion analysis is especially investigated. The unifying thread is the design of active remote monitoring systems requiring efficient methodological tools for dynamic vision, image sequence communication and local control through active vision, or remote control by man-machine interaction.
Research themes:
… Motion analysis for dynamic vision.
… Multiscale statistical models.
… Active vision based on visual-servoing.
… Image sequence coding.
… Compression and interpolation.
2.1.2.8. SIAMES (Image Synthesis, Animation, Modeling and Simulation)
This project considers all methods needed for producing sequences of computer-generated images. There are three key tracks:
- computer graphics, where we are seeking to define algorithms,
- simulation, with the aim of matching our algorithms against numerical values measured on a real site,
- the system organization to process "life-size" cases and validate our approach by implementing them.
Research themes:
… Modeling complex scenes (3D) and especially the shape of objets in terms of constraints and properties.
… Simulation of lighting: introduction of lighting models based on physics.
… Simulation of physical systems: study of computing diagrams needed to produce the equations controlling these systems.
… Parallel algorithms: research into reducing the complexity of sequential algorithms and the study of parallel schemes for these algorithms.
2.1.2.9. PRISME (Geometry, Algorithms and Robotics)
This project is dedicated to developing and studying geometric structures and algorithms, and their implementation and applications. Applications involved are mainly but not exclusively derived from robotics.
Research themes
… Design and analysis of geometric algorithms:
- General algorithmic techniques: randomized, adaptive algorithms
- Geometric optimization
- Voronoï diagrams
… Programming of geometric algorithms:
- Robust algorithms
- Development of the CGAL library
… Geometric modeling of shapes in medicine and geology.
… Robot motion planning.
… Mechanism modeling, especially parallel robots, using geometric and algebraic approaches.
… Applications in the following fields: positioning for cutting, satellite equipping, medicine, computer-aided design.
2.1.3. The Engineering Science Consortium initiative
In November 1995, the Centre National de la Recherche Scientifique (CNRS) considering both national and international aspects proposed a joint initiative to the French institutions engaged in Engineering Science R&D.
Streaming from this initial step, CNRS and CEA were entrusted by the Ministry of Science and Technology (MENESR) together with organizations such as CEMAGREF, INRA, INRIA, ONERA,... to define and to set a consortium that could encompass the development of technology subjects central in todays Engineering Science.
Important factors have been considered to shape the Engineering Science Consortium:
The general consortium objective consists to blend those factors in a set of well-structured actions with a special emphasis in R&D projects structured around the perspectives and the technical requirements of one or several industrial partners.
Thus, the Consortium provides the appropriate framework for a set of medium term projects combining in a flexible way the know-how and the expertise of generic research institutions such as CNRS or INRIA with the domain-oriented ones such as CEA, CEMAGREF, INRA, ONERA...as well as the network of Technical Industrial Centers and R&D departments of industrial companies.
The general structure and organization of the consortium has been completed. It includes a council of partners representatives, a Directorate of three members and an Advisory Committee comprizing Industry (10) and Ministeries (5) representatives.
The Consortium will undoubtedly host projects related to the field of IARP. They will be open to international cooperation, in particular at the European level.
2.2. GENERAL BASIC RESEARCH AND ENABLING TECHNOLOGIES PROJECTS AND PROGRAMMES
2/ERB4050PL920998 - HEROS NETWORK: Hazardous Environment ROboticS
Coordinator : LIFIA-INRIA (F)
Participants : LAAS-CNRS (F)
SCUOLA SUPERIORE S ANNA (I)
Univ. of OXFORD (UK)
Univ.of LEUVEN (B)
Univ. of KARLSRUHE (G)
Inst. de CIBERNETICA (SP)
Start date : September 1, 1993 - End date : December 31, 1996.
HEROS is a Network which addresses the general problem of using and controlling robots in hazardous environments.
The objectives of HEROS are: to develop the cooperation between European institutions working in Robotics, to facilitate the exchange of researchers between these institutions, to develop the required human resources, to increase the competitiveness of the EU in
the Robotics field, and to favour technology transfer and dissemination of information.
The HEROS network brings together eight European Research groups selected for their scientific and technological know-how covering a set of particular key domains: task-level programming and planning, sensor-based modelling and model-based sensing, and sensor-based task execution and control.
The major scientific achievements obtained are:
- The integration of planning, sensing, visual servoing, and control for automatic grasping in a partly known environment.
- The integration of vision, fine-motion planning, and force-controlled to achieve an insertion in the presence of uncertainties.
- Motion planning for mobile robots with trailers and extension of the elastic band method to non holonomic systems.
- A new control scheme for dextrous hands.
INTAS 94-2666 Project MORES : Advanced Mobile Robots for Specific and Difficult Environments
Duration : January 1995-December 1996
Coordinator : LAAS-CNRS, Toulouse (F)
Partners : IPR, Karlsruhe (G)
Keldysh Institute, RAS (RU)
Institute of Mechanics, RAS (RU)
Institute for Problems in Mechanics,
RAS (RU)
Institute for Systems Analysis, RAS (RU)
HEUDIASYC, Compiègne (F)
LAN, Nantes (F)
LIFIA, Grenoble (F)
LRP, Paris (F)
DLR, Institute for Robotics,
Wessling (G)
Project objectives :
There is an increasing need for machines that will carry out tasks in working environments requiring mobility for their execution. Among the broad variety of related application domains, very significant cases are: planetary exploration, underwater work, intervention in hazardous sites, mining, shipbuilding, waste handling, nuclear sites maintenance, building maintenance, fire fighting, forestry, etc., which all imply remote operation from a distant site.
With the various fields of applications given above in view, the research programme covers four important subjects:
The project was successfully completed by June
1997.
COPERNICUS Project : Multi-Agent Robot Systems for Industrial Applications in the Transport Domain
Coordinator : Inst. for Real-Time Computer Systems & Robotics,
Univ. of Karlsruhe, Germany
Partners: INRIA, Grenoble, France
Techn. Univ. of Budapest, Hungary
Techn. Univ. of Poznan, Poland
Russian Academy of Science, St Petersburg, Russia Belarussian State Univ. Belarussia
UFA Aviation Techn. University Mercedes-Benz AG, Stuttgart, Germany
Duration of the project : 01.01.1997 - 31.12.1998 (24 months)
Description of the project:
The development of an open and flexible multi-agent mobile robot system with a distributed control architecture and intelligent information processing for various industrial applications in the transport domain is proposed. Due to industrial need to create Intelligent Manufacturing Systems (IMS) that can automatically perform complex technological tasks the proposed research efforts concentrate especially on the development of intelligent multi-agent mobile robot systems for autonomous execution of transport operations in a factory.
Within the project a generic approach for designing multi-agent mobile robot transport systems for IMS will be devised based on a multi-level control system architecture with
intelligent features including the combination of neural network and fuzzy logic approaches.
The resulting multi-agent robot systems will have the capability of autonomously controlling and carrying out various monotonous technological transport operations.
BR 8173 : BARMINT (BASIC RESEARCH FOR MICROSYSTEMS INTEGRATION)
Duration : April 1994 - April 1997
Coordinator : LAAS-CNRS (F)
Participants : Institut National Polytechnique de Grenoble (F)
Technische Hochschule Darmstadt (G)
National Microelec. Research Center (IR)
Centro National De Microelect. (SP)
Univ. Karlsruhe (G)
Subcontractors : Tech. Univ. of Lodz (PL)
Tech. Univ. of Budapest (H)
BARMINT aims to identify and develop the design methodologies, tools and technologies necessary for designing and fabricating Silicon based microsystems in monolithic and multichip conditioning.
A common demonstrator, baseline for the work-packages of this interdisciplinary project, tackles the central problem of compatibility and clusters the difficulties to be encountered in future industrial microsystems. It will associate optical, mechanical and chemical functionalities to propose solutions to the technological bottlenecks in compatibility.
Beyond the technological feasibility demonstration of Silicon based complete compact microsystems, the essential results of the project will concern basic methodologies (CAD tools for microsystems mainly in functional, 3D thermo-electrical and thermo-mechanical simulations) and processes (Micro-devices and related processes and new developments in 3D multichip packaging taking into account mechanical and chemical constraints) for top-down microsystem design.
In the current phase BARMINT has already obtained important results related to two basic microtechnology processes : packaging and thermo-machanical constrains in stacked multichip systems. These results have been successfully applied in various demonstrators :
- multisensor and multiactuator components
- integrated systems including a feasibility prototype of a micropump.
FMRX CT 96 0052 - The TMR Network: SMART II -Semi-autonomous Monitoring And Robotics Technology-
The SMART network brings together ten European laboratories and two start-up companies active in the use of computer vision and mobile robotics for video surveillance and monitoring using both fixed and mobile camera.
The SMART network makes possible exchanges of doctoral and post-doctoral researchers between EU countries, and between universities and SME's.
Activities within SMART concern traditional security concerns such as visual tele-surveillance and autonomous mobility for monitoring and exploration of hazardous areas such as underwater and nuclear environments.
Network partners are applying technologies developed under SMART to new applications areas such as robots for detecting land mines, mobile service robots for assisting handicapped patients, video-communications, and computer assisted remote teaching.
Recent technical achievements include the demonstration of a real time system for integrating and controlling visual behaviors for surveillance and for navigation by INPG, Aalborg University, University of Genoa and the SME AITEK. A real time system for tracking vehicles and people in real time has recently been demonstrated by University of Leeds and University of Reading. Many of the partners have developed active robotic heads, including a commercial head marketed by RobotSoft, and a head widely reproduced by research labs designed by Aalborg University.
A key partner is the SME Society RobotSoft in Bidart, France (near Biarritz). RobotSoft is the second largest manufacturer of mobile robots for experimental applications, worldwide. Many of the SMART partners use RobotSoft mobile robots. The SMART network makes it possible for partners to send researchers to work with RobotSoft in the development of new technologies for mobility and surveillance.
The SMART network began in 1993 with a 30 month contract period and ten partners, two subcontractors, and several industrial affiliates spread over 6 EU countries. The network's goal is to aid mobility of scientific researchers, technology transfer and dissemination of information. The network organizes a major annual international symposium (SIRS), and semi-annual technical workshops. During SMART-I, 56 young European researchers spent a total of 204 man-months working in laboratories outside of their home country. A new contract for SMART II has begun in January 1997 with an expanded budget for a period of 36 Months. Roughly 400 man-months of funding are available for transnational visits of young scientists and engineers to work in network laboratories in France, Italy, Denmark, Portugal, the UK, and Ireland.
The SMART network is coordinated by Professor James L. Crowley of the INP Grenoble. For additional information, consult the web site at http://pandora.imag.fr or contact professor Crowley by electronic mail at (jlc@imag.fr).
2.2.2. Microrobotics and Microsystems
2.2.2.1. French initiatives on microsystems technology
France, as many other countries (Germany, Japan, USA,...) recognizes the importance and the possibility of a major advance in Microsystems Technology that will open a broad spectrum of novel applications with considerable technical and economic impact.
Since 1990 France has been conducting studies on microsystems through national organizations such as CNRS and CEA and has organized, under the authority of Science and Technology governmental institutions, R&D cooperative projects associating industrial companies and laboratories.
Appropriate initiatives are also taken by France to include these R&D efforts in the framework of European institutions as well as to establish other international cooperations (cf. section 2.2.2.3).
In 1993, building up on previous work in the domain of smart integrated microsensors the Ministery of Science and Technology decided to launch a French Initiative in the field of Microsystems Technology with the aim to develop a French coherent R&D effort in the field.
In coherence with this initiative, CNRS is engaged, often with a leadership role, in three sorts of national actions:
i) A federative project in Microrobotics and Microsystem (cf. section 2.1.1.10),
ii) The creation of two microtechnology centers:
. the Microtechnique Institute of Franche Comté (IMFC), located in Besançon (France), a public research center comprising 120 scientists and technical staff,
. the development of Microtechnology cooperative facilities at LAAS, Toulouse
iii) A National Research Group on Microsystems with a very important twofolded role :
. to foster basic research in the field and to provide the framework for scientific exchanges and cooperation,
. to provide a coordination frame and a general perspective to application domains and related actions including the two above mentioned (i), (ii).
2.2.2.2. CNRS Research Group on Microsystems
The purpose of the Research Group on Microsystems is to promote coherent and cooperative microsystems-oriented research work.
The general objectives of the Research Group programme are the following:
- study of new processes specially targeted for microsystems: micromachining and three dimensional pattern generation,
- development of 2D and 3D assembly techniques,
- design, elaboration and characterization of new non electronic devices: sensors, motors, actuators, microresonators, microswitches, microoptical components,
- development of CAD tools for microsystems integration.
The main organization feature of the Research Group on Microsystems associates task forces to four demonstrators. The following task forces/demonstrators have been created with 18 CNRS laboratories:
- Silicon Integrated microprobe for medical applications,
- Electromagnetic and/or optical microsystems using electrostatic deflection,
- Interfacing and multiplexing for an optical array sensor,
- Microelectromechanical devices for microrobotics.
In 1997, the CNRS decided to extend this activity into a more general programme enhancing pluridisciplinarity. New projects start now with applications in chemistry, biology, nanofabrications, etc. Three types of objectives are supported by the programme:
This programme is created for four years. It is sponsored together by the CNRS and the Ministry of Research.
2.2.2.3. International cooperation
EUROPE
NEXUS
Specially worth to mention is NEXUS, the Network of Excellence in Multifunctional Microsystems, sponsored by the European Communities in the frame of ESPRIT.
NEXUS is a joint European effort to prepare European industry for the microsystem world market by coordinating European R&D in Microsystem Technology (MST).
The objective of NEXUS is to provide a communication network between its partners to exchange know-how and technological skills, to give strategic guidance for basic R&D and to disseminate knowledge on MST.
NEXUS ended in December 1995. NEXUS II project has a shifted focus on information dissemination and industrial-oriented objectives.
Among the new features, there is MST/MEMS on the net.
EMSTO (European MicroSystem Technology Online) is a World Wide Web service established under the auspices of NEXUS (the Network of Excellence on Microsystem Technologies). EMSTO provides a comprehensive information service to the European Industrial and academic community on microsystem (MST) or microelectromechanical system (MEMS) technologies.
The EMSTO mission is to facilitate the exchange of MST/MEMS information so as to promote the technology particularly to the European industrial community.
The service is managed through a consortium of companies and academic institutes: LAAS-CNRS (F), MSTB Ltd (UK), ECAD (UK) and VDI/VDE-IT (D).
JAPAN
The CNRS (France) and the Institute of Industrial Sciences (IIS) of The University of Tokyo signed in March 1995 a joint research agreement in the field of Microtechnologies and Microsystems creating the "Laboratory for Integrated MicroMechatronic Systems" (LIMMS). Within the LIMMS, cooperative research work is carried out in the field of Micromachatronics at the focal center of Roppongi, Tokyo, by:
. Institut d'Electronique et Microélectronique du Nord (Lille),
. Institut des Microtechniques de Franche Comté (Besançon),
. Laboratoire d'Analyse et d'Architecture des Systèmes (Toulouse).
The technical topics of the joint research include the field of micro-electromechanical systems that are directly related to the microrobotics domain.
In 1998, it will be extended to other laboratories and to LETI-CEA. A new research theme will be developed on micro-nanorobotics for medicine and biology.
Medical robotics together with related image and information processing subjects constitutes a fast growing domain highly relevant to IARP objectives in which several French groups belonging to CNRS, INRIA, INSERM and Universities are successfully engaged both through local projects and cooperative research efforts.
Fostered by wide-spread interest, national and international initiatives are currently engaged. Among them, we shall specially mention:
… The project organized around the TIMC/IMAG (Grenoble) laboratory in the frame of the CNRS strategic action (cf. section 2.1.1.6.),
… The EPIDAURE Project carried on by INRIA and centered in the Sophia-Antipolis site (cf. section 2.1.2.3.).
IGOS: Image Guided Orthopedic Surgery (Telematics program)
This project focuses on the clinical evaluation of a large range of techniques for image guided surgery in orthopedics and the definition of guidelines for the selection of techniques according to a clinical target. Robotized and non robotized techniques are clinically evaluated within IGOS.
Coordinator: CHU Grenoble; Educational, Clinical and Industrial partners in
France, England, Italy, and Germany.
CRIGOS: Compact Robot for Image Guided Orthopedic Surgery (Biomed Program)
This project focuses on the development of a new robot for Image Guided Orthopedic Surgery. A compact robot based on a Stewart platform should provide more rigidity and safety than a conventional robot (Robodoc for instance) and should be more adapted to the conditions of the operating room. A system based on such a robot is developed and will be evaluated in the clinical environment.
Coordinator: Helmholtz Institute, Aachen, DE; Educational, Clinical and
Industrial partners in France, England, Italy, Germany, the Netherlands
MITS: Mechatronic Invasive Tools for SUrgery (Biomed concerted action)
This project is intended to create links between medical, industrial and technical research teams throughout Europe with interests in the area of medical mechatronics. It should discuss the requirements of such tools and result in the definition different standards (safety is one).
Coordinator: AMARC, Bristol, GB; Educational, Clinical and Industrial partners in France, England, Italy, Germany, Greece, Austria, Ireland, Belgium
2.4.1. Status and perspectives
France has a long standing record of activity in the domain related to Robotics in Agriculture. This activity was initiated by CEMAGREF and continuous R&D efforts have been pursued under its leading role. CNRS laboratories began to work jointly in the domain as of 1980. CEMAGREF launched in 1987 the pioneering MAGALI project for apple picking.
Autonomous mobile robotics research programmes for agricultural equipment will progressively come to an end. Expertise developed in CEMAGREF on these programmes in the past ten years (i.e. metrology, real-time sensors for environment perception and for quality control, multisensor fusion ; information processing ; human knowledge and know-how modeling and representation, artificial intelligence, control/command) are applied to new research fields: equipment and information systems for a clean and precise agriculture on one hand, and monitoring and control of the food products quality along the transformation processes on the other hand. The final aim is the design of information and Decision Support Systems which put the operators back at the center of the process.
In the reminding part of this section, we describe the most important international cooperative projects in which France plays a meaningful role.
EU 176: CITRUS-ROBOT FOR CITRUS HARVESTING AND HANDLING
Objectives: Development of a robot prototype for detection, collection and manipulation of citrus fruits.
Duration: 5 years, starting in 1992
Participants: PELLENC & MOTTE, SAGEM, CEMAGREF,
IVIA (F)
INDUSTRIAS ALBAJAR, CASA (SP)
EU 331: ROSAL
Objectives : Development of robots for rose bush handling and grafting.
Three different concepts are considered : digging, selection and grafting.
Duration: 54 months, starting in 1993
Participants: MEILLAND, PELLENC & MOTTE,
CEMAGREF, IVIA, ITMI, INRIA (F)
Universal Plantas, IASA, INELCOM,
IASA, INELCOM, Universidad Tecnica de Valencia (SP)
EU 705: CIMIS - COMPUTER INTEGRATED MILKING SYSTEM
Objectives: Development of a robot for automatic milking of cows.
Duration: 3,5 years, started in 1992
Participants: CEMAGREF, SAGEM, DIABOLO SANUS (F)
PROLION (NL)
ESPRIT IV project 9230 - SHIVA
Development of sensor and handling systems for automatic on-line sorting and quality evaluation and packing of fresh products.
Objectives: Robotized inspection of fruits.
Duration: 3 years, starting in 1994.
Participants: PELLENC S.A. (F)
CEMAGREF (F)
Food Machinery Espanola S.A., IVIA (SP)
SYNTAX FACTORY AUTOMATION (UK)
The overall goal of the project is to design and implement an integrated system including:
- new sensor systems for on line complete quality control
- robot cells equipped with vision and sensor systems for optimum sorting and packing
- flexible fruit box handling systems, optimized for different products and containers
- CIM management of the packing, using advanced control system architecture.
2.4.4. Specific European Projects
BRPR-CT9: CARETAKER - ENVIRONMENTAL FRIENDLY FOREST MACHINE TECHNOLOGY
Duration: 3 years, starting in 1996
Participants: CEMAGREF (F)
TIMBERJACK OY (FIN)
SAUER-SUNDSTRAND Gmbh, Univ. Paderborn (D)
STORA SKOB AB (S)
Elaboration of the specifications and requirements of future forest skidding machines
Fully hydrostatic translation system coupled with spinning control, digital command crane, intelligent man-machine interface, Hydraulic and electronic architecture (distributed CAN system). Design and development of a prototype.
AIR-31072: OBJECTIVE PLANT QUALITY MEASUREMENT BY IMAGE PROCESSING
Duration: 3 years, starting in 1994
Participants: CEMAGREF (F)
Univ. Agr. Wageningen, PBN (NL)
Royal Veterinary and Agri. Univ. Inst. of Plant and
Soil Sciences (DK)
Ultimately, image processing will give an objective assessment of pot plants and cut flowers for 1) production improvement (growth monitoring and management) and 2) setting up European quality standards; This project combines vision technologies, neural networks, fuzzy logic and use of genetic algorithms.
PL96-199: MACA - MEAT AUTOMATION CONCERTED ACTION
Duration: 2 years, starting in 1997
Participants: CEMAGREF (F)
AMARC Bristol University (UK)
Plus several other participants from 15 countries
The aim of this concerted action is to gather research teams, the meat processing industry and equipment manufacturers to identify the technology transfer possibilities from other sectors, and the priority R and D topics for meat processing automation.
AIR21204: IN-SPACE - REDUCED N-FERTILIZER INPUT BY AN INTEGRATED, LOCATION SPECIFIC MONITORING SUPPORT AND APPLICATION SYSTEM
Duration: 3,5 years, starting in 1994
Participants: CEMAGREF (F)
Univ. Agr. Wageningen (NL)
Silsoe Research Inst. (UK)
Univ. Catho. Leuven (B)
Expected results: 1) in-field sampling methods 2) GPS localization systems and data collection software 3) comparison of quick nitrogen soil content analysis techniques 4) dynamic crops reflectance measurement system 5) crop yield on-line measurement system 6) data exchange software with a GIS 7) automated fertilizer application system regulated accorded to the position.
2.5.1. CEA - TELEOPERATION AND ROBOTICS DEPARTMENT
The Teleoperation and Robotics Department (STR) belongs to the Advanced Technologies Division (DTA) of the French Atomic Energy Commission (CEA). The department is in charge of the Research and Development in the field of Teleoperation and Robotics, with the aim to mobilize the corresponding CEA resources towards industrial application sectors which concern computer-assisted teleoperation, advanced manipulators, mobile robots and associated systems (man-machine interface, environment modeling, on-board electronics...). The department is mainly involved in nuclear applications : decommissioning, reprocessing, inspection... Technology developments are also used for other " hostile " applications (as underwater, military) and services (medical). The STR has a staff of about 70 persons (engineers and technicians). A team is specially in charge of testing robotic systems in conditions representative of real tasks to validate the applicability of the developments.
Hereafter is given a description of the main nuclear-environment oriented developments:
- BP 250 is redundant C7 of dexterous manipulator for maintenance in refueling installations. The payload of the robot is 25 kg for a weight of 75 kg. The robot is force feedback controlled for dexterous teleoperation.
- RD 500 manipulator is a robust, watertight robot able to handle payloads of 500 N. It has been qualified, in the frame of the decommissioning programme of the DGXII. A two arms system is tested to demonstrate teleoperability of maintenance tasks in the frame of the ITER project (fusion).
- CENTAURE is a watertight mobile robot fitted with a telemanipulator and designed for inspection in a nuclear plant in accidental conditions.
- MESSINA is a multi-body vehicle able to carry a heavy load in a very cluttered environment (including stairs). The operator pilots the first module, and the other modules follow automatically the same path (TELEMAN project).
- SHERPA is a six-legged mobile robot able to transport a payload of 150 kg (TELEMAN project).
- PML is an innovative long reach carrier concept of self constructed carrier allowing accessibility to a very large range of hot cells (TELEMAN project).
- TAO 2000 is a computer aided teleoperation control system, providing an extended set of manual, shared and automatic modes, including hybrid force-position and sensor-referenced controls. TAO 2000 has been connected to various kind of manipulators : MA23 (electric master-slave telemanipulator), RD500 (electric telerobot for dismantling applications), RX90 (Industrial manufacturing robot), MAESTRO (Hydraulic manipulator).
- PYRAMIDE is a 3D interactive graphic system used to monitor and control the tasks performed by telemanipulators as well as mobile robots.
- Graphic man-machine interface are developed for the control and supervision of the robotics systems. They are based on virtual reality technics.
2.5.2. EDF: Robots and remote-control research and development
EDF intends keeping a Research & Development activity in Robotics and remote control operations for several reasons:
- on the one hand, to have some knowledge and an ability of assessing the technologies available in France and abroad, a standing action of technological survey being carried out on the subject,
- on the other hand, to carry out research and development actions in a certain number of fields considered significant so as to have a technical expert assessment available, and to participate to overcome certain technical bottlenecks ; these actions are either achieved on technological components (or technological blocks), or in selected subject themes,
- and lastly, to stress the interest of a system/process with laboratory experiments, and the possible improvements.
Part of this process is achieved within a pluriannual Research and Development Project: START (Projet de Système de Téléintervention Avancé Robotisé Transportable, or "Advanced Transposable Robotized Remote-Control operations Project") for nuclear repair applications. Part of this R&D also concerns the use of remote control and robotics for non nuclear EDF applications (interventions on energized electric lines, robotized repair of hydraulic turbines bladings, etc.).
Among the research themes and the various technological components under study, the following seems specially worth to mention:
- Handling systems control/command: some work is being done on manipulator control/command: this includes redundant manipulator control command (EDF has several 7 d.o.f. light weight manipulators), vision based control, and mixed force-position control. Robust software is being developed in order to be used in industrial environment.
- Stereovision: Following an assessment of the various technologies and ergonomics carried out in collaboration with the French Ecole des Mines de Paris, the industrial prototype of a stereovision system was realized jointly with the LTD company in Montpellier, France. The system reveals many advantages as compared to the existing systems. Any store-bought camera can be adapted to it, and can reproduce the relief at a 25-image-per-second rate, on a scintillationless screen.
- Mobile robots: EDF is the coordinator of the TELEMAN project IMPACT which ended in 1996. It has also developed, associated with CEA and COGEMA, mobile teleoperated vehicles for emergency situation. This includes:
. Mobile teleoperated vehicles for inside inspection (ERII)
. Radio controlled vehicles for outside inspection (VERI 2A, VERI 2B, ERASE)
. Radio controlled civil engineering services engines (chain loaders, excavator, ...)
- Robotized or remote control maintenance of future standardized plants: Experience feedback has shown the difficulties in relying on robots or remote control equipment when interventions have not been considered in the original design of the plants. Being the industrial architect of its own plant, EDF takes part in designing the future plants so as, subsequently, to facilitate the implementation of robotized or remote control systems. The foreseen arrangements concern mainly high capacity communication networks, easy ways to achieve absolute localization in space, and a less constrained spatial environment.
- 3D laser sensor 501SIC : (Laser Mapping System) : the Laser Mapping system is a laser scanner designed for 2 m to 30 m range. It is used to generate as built CAD Models of Nuclear component from Nuclear Plants. An associated software, 3D ipsos, is used to generate CAD models from the 3D points.
TELEMAN is a programme of research for the European Atomic Energy Community in the field of remote handling in hazardous or disordered nuclear environments. The programme running since July 1989, is managed by the Commission of the European Communities Directorate General XII for Science, Research and Development in close collaboration with the TELEMAN Users Group. It will continue till 1996.
The research in the TELEMAN programme is aimed towards developing tele-operated machines which could take over tasks currently performed by human operators, thereby reducing their exposure to radiation.
F12T - 0028 - A LIGHT MOBILE TRACKED MACHINE WITH ADVANCED SENSING, WORLD MODELING AND COMMUNICATION - TELEMAN "IMPACT"
Start date: March 1993 - Duration : 36 months
Participants: EDF (F)
INITEC (SP)
KERNTECHNISCHE HILFSDIENST GmbH (G)
TNO TPD (NL)
Univ. COLLEGE LONDON (UK)
KENTREE (IRL)
LOUGHBOROUGH Univ. of Techn. (UK)
FRAUNHOFER GESELLSCHAFT-IPA (G)
Univ. Libre BRUSSELS (B)
CIEMAT (SP)
CEA/LETI (F)
PIAP (P)
Objectives: The overall objective of the IMPACT project to develop a light mobile robot which can be quickly reconfigured, with different sensors and/or communications, for a variety of inspection tasks within nuclear plants. These interventions are to cover the range of normal, incident and accident situations and aim at surveillance (or data collecting) missions.
Among the sensors to be integrated within the project, let us quote the sensors used in 3D environmental reconstruction, using either the flight-time laser technique (the French Atomic Energy Commission Alis camera), or the stereovideogrammetry (the University College of London HAZMAP system). A radiation camera, giving a radiation "image" is also currently under study (the Spanish company CIEMAT). All these sensors have been implemented in 1994 in a Spanish plant, to reconstruct the plant's 3D environments on the one hand, and to establish a radiation cartography of certain areas on the other hand.
The final demonstrations of the project have been done in a Russian nuclear plant, in Smolensk ; the IM robot worked in cooperation with a Russian robot, to built a protection wall. Another demo has been done in the East-Germany power plant of Greiswald, with the robotized version of IMP.
2.6. HUMAN ASSISTANCE ROBOTIC SYSTEMS
Following the first five years of the PREDIT program (Innovative technologies R&D program for ground transportation), a second large scale phase for PREDIT has been implemented by the Ministry of Equipment, Transportation and Tourism, in cooperation with the Ministry of Research Technology and two major national R&D agencies. The program aims to pursue a major R&D cooperative effort implying industry, R&D national institutions and university laboratories.
The scope of the program covers of course many aspects related to ground transportation that do not pertain to the field of IARP.
Three important objectives are directly related to IARP:
- improved safety features for vehicles and vehicle roads and ground installations,
- system development for improved vehicle performances in relation to environment pollution (noise, exhausts,...),
- reduced energy consomption.
The Praxitele programme which was started in May 1993 is carried on by a consortium of industrial partners : CGEA, Renault, Dassault A.T., EDF and two national research institutes : INRIA and INRETS.
Praxitele is a personal public transport system with Machine Intelligence characteristics which avoids the drawbacks of the automobile - urban congestion and pollution - and offers its main convenience - the freedom of movement -. It is based on small electric vehicles available at any time on specific areas. Clean and silent, they are environmentally friendly and save on energy. Praxitele allows the individual to drive freely and easily around the towns and suburbs, without having to worry about parking and maintenance.
Praxitele is an extension and a complement to traditional public transport when and where they are not efficient. It is designed for short local trips or schedules where there is low and medium demand, especially during off-peak hours. It also offers very convenient access to mass transportation. Users should easily move from low density to high density zones.
IFREMER (Institut Français de Recherche pour l'Exploration de la Mer) considers subsea robotics as one of the major Technologies for present and future developments in Ocean Engineering. Therefore, it is committed to sustain a four-folded effort :
- a robotics laboratory has been set in Toulon,
- IFREMER has created under its leadership an R&D consortium that will carry on projects in the domain of underwater robotics covering application domains ranging from off-shore building and maintenance to science oriented engineering.
Besides IFREMER, the consortium, Robotique Sous-Marine Méditerranée (RSM) includes:
• INRIA (IRISA, Sophia),
• eleven industrial companies and organizations,
In 1997, to cope with a wide set of applications, the RSM opens its activity domain to include the general area of robotics in hostile environment ; the new name of this association is GRSM South Mediterranean Robotics Group.
ROV 6000 VICTOR - The new 6000 m teleoperated ROV VICTOR building is finished and the first at sea tests are forecast in September 1997.
This new operational tool for IFREMER shows the robotics orientations chosen by the Institute.
France will host the OCEANS'98 conference in Nice.
AIOLI is a research project involving CNRS, IFREMER, THOMSON, INRIA and CYBERNETIX, funded by the Regional Council of Provence in the framework of the RSM consortium. The objective of the project is to develop a regional testbed for research on subsea robotics including an experimental ROV with telemanipulation. Telemanipulation devices and new correlation log were developed and tested in 1997.
2.7.3. European Union projects
MAST II - NEW METHODS FOR DEEP SEA INTERVENTION ON FUTURE BENTHIC LABORATORIES : ANALYSIS, DEVELOPMENT, ENGINEERING AND TESTING
Participants: IFREMER (F)
GEOMAR (D)
THETIS (D)
Instituto Superior Tecnico (P)
The project deals with the operation of future benthic laboratories and stations. Four different concepts are being studied: a mobile hook, a dedicated ROV, a scientific ROV and a free swimming vehicle. For each concept, engineering studies are conducted ; comparative studies of dynamic behavior and acoustic performances are carried out. In some cases, equipment or mock-ups will be realized and tested at sea.
Amongst the tools at the disposal of the scientific community, benthic laboratories and benthic stations will experience and increasing interest on account of the missions they will permit: observation, measurement and long term experimentation completed on the sea floor.
The general objective of the project is to investigate new methods for deep sea intervention that could be used in the future to deploy, supply, operate and recover benthic laboratories.
Four concepts are investigated:
The optimum scope of application of each system will be determined according to:
- the benthic lab characteristics (size, weight, life duration),
- the vessel characteristics (positioning, weight, handling capability, heave compensation),
- the applications (lab handling, labs interconnection, energy source recharging, data recovery).
Esprit BR 8972 - UNION - UNderwater Intelligent Operation and Navigation
Duration: 24 months
Start date: June 1994
Participants: IFREMER (F)
INRIA (F)
LIRMM (F)
UPC (SP)
TU-MUNCHEN LBM (D)
Heriot WATT UNIV. (UK)
OXFORD UNIVERSITY (UK)
The central aim of the UNION project is to develop methods for increasing the autonomy and intelligence of unmanned underwater vehicles (UUVs). The project has one central theme which concentrates on the integration of heterogeneous and traditionally disparate processing, actuation and sensing subsystems in a complex and hostile environment. This theme can be first focused on :
i) Coordinated control strategies for combined vehicle-manipulator systems:
In underwater vehicle applications it is very difficult to decouple the control of vehicle motion from the control of manipulation internal forces/couples when involved in the execution of typical tasks. This is because the forces exerted by the manipulator are reflected through the vehicle resulting in positional changes and consequent lack of manipulation accuracy. This situation is currently overcome by clamping the vehicle to the structure on which work is being conducted. However this limits both the variety of tasks that can be performed by the vehicle and the nature of the environment in which the combined vehicle-manipulator system can operate. The aim is to look in detail at various aspects of this coordinated control problem and to develop new systems capable of combined manipulator-vehicle positioning and control.
ii) Integration of data from different underwater sensor systems to provide robust and reliable information on vehicle location and environment description information:
Underwater sensors differ markedly from sensors used in land-based autonomous systems. In particular the basic physics of underwater sensors are less-well understood, and noise levels false alarm rates are much higher than in comparable land-based systems. The aim is to apply a number of new data fusion techniques developed for land-based applications to model and integrate data from a number of different sonar in particular laser-ranging and vision systems available within the consortium.
The project, completed by the end of 1996, was highly successful and several functions are under valorization on operational systems. A new offspring project dedicated to arm/vehicle stabilization and coordination will be launched with industrial partners.
Field robotics represents today a cutting-edge research stream in the domain of autonomous mobile robots, i.e. Intelligent Machines possessing decisional and operation al autonomy. There is a rich spectrum of application domains, most of them implying highly demanding technical advances in subjects such as automated locomotion, 3-D perception and modelling, on-board autonomy, and task -level teleprogramming.
France has been very active in three application fields implying natural terrain non-cooperative environments. This is the case of open-terrain surveillance (as well as general defence oriented applications), planet exploration and ground intervention in remote and highly hard places such as Antarctica.
The following sections highlight the on-going programmes lead by the French Space Agency (CNES), and several laboratories and R&D organizations
Various Lunar and Martian missions have been under consideration with a main thrust on a technological mission on the Moon in 2002 (Lunar European Demonstration Approach -LEDA-). The mission aims to demonstrate the feasibility and operational capacities of an autonomous rover covering hundreds of kilometers in difficult terrains.
The framework of this activity has been actively pursued in a three folded way by CNES:
In this context, the consortium of RISP laboratories that since October 1989 pursued a broad spectrum of research activities scientifically related to CNES oriented projects, concluded his activity on September 1996.
Streaming from the work carried out by RISP, three on-going developments deserve a special mention. They are by chronological order :
Streaming from the ESA initiative for a possible major international effort to carry on ambitious Lunar missions, and in the frame of the foreseen project LEDA, the French Space Agency has implemented the EVE experiments in the GEROMS testbed.
The EVE vehicle, based on a Marsokhod-like platform, has been integrated at CNES to validate the following operation modes:
- Autonomous motion based on stereovision perception and autonomous path planning on board.
- Teleoperation in Adquired Synthetic Environment (TESA) with real time Digital Terrain Model production on board.
During the EVE tests, all algorithms were running on the on-board computer, as they would do during a planetary mission. The tests, performed during six months, have proven the feasibility of both modes, and have been used to elaborate higher level strategies (for example, the selection of the perception orientation) that will be implemented on the IARES vehicle. A continuously planning mode, where the vehicle performs perception and planning during motion, has also been tested on EVE.
EVE is currently operated, as a common research platform, by LAAS and CNES.
2.8.3 GEROMS (Groupement pour les Essais en RObotique Mobile Spatiale)
GEROMS was created in August 1992 following a decision by the CNES/CST (Centre Spatial de Toulouse), the ONERA/CERT and the CNRS/LAAS to pool their individual competence and develop a common test facility for mobile robots in order to promote a centre of expertise unique in Europe for the testing of autonomous vehicles.
GEROMS objectives are the experimentation, verification and validation of design concepts, the unit tests of subsystems fitted on a test vehicle, the development of complete systems with a view to integration and final acceptance, and finally, the experimentation and testing of mobile robot demonstrators in a representative context which considers mission peculiarities in terms of the environment.
In order to achieve these objectives, GEROMS has a main test site which consists of both an instrument-equipped automated test vehicle, and a calibrated test site including obstacles commonly met by robots during outdoor operation. This test site may be broken down into a "Moon" area, an "Earth" area and a "Mars" area.
The main test site is also equipped with the following:
- measurement facilities designed to firstly calibrate and characterize the site and then to assess the performance of both onboard subsystems and the robots themselves
- software development equipment, which consists of workstations and analysis tools to debug and modify onboard software, whether on the main test site or during campaigns,
- computing facilities used to interface with and simulate the environment of systems or subsystems being tested,
- simple locomotion facilities (not autonomous, versatile, etc.) for subsystem testing,
- infrastructure providing the necessary support and resources (electrical power supply, air conditioning, etc.) during campaigns.
The test facilities and part of the infrastructure are "transportable" so that it is possible to move all such resources to an external test site considered more representative of the final operating environment (in situ tests).
Following the research programme carried out by CNES in association with RISP, it was decided to build a demonstration rover to validate the robotics functionalities and to experimentally demonstrate the degree of operational and decisional autonomy offered by concepts developed since 1989. The Eureka pan-European structure appeared to offer a logical continuation to the French studies while also seeking to federate and transfer to the field of space the expertise and the achievements generated by other programmes in mobile robotics in which both industry and research participate. The reverse, i.e. the spin-off from space to industry is also a meaningful and important objective.
IARES associates 9 partners from 4 countries:
France: ALCATEL ESPACE
CYBERNETIX
ITMI
MATRA MARCONI SPACE
SAGEM
RISP
Spain: IKERLAN
Hungary: KFKI/RMKI
Russia: VNIITRANSMASH
RISP, IKERLAN, KFKI and VNIITRANSMASH represent several organizations or sub-partners.
The objective of the IARES project is to draw up a detailed definition of the sub-systems needed to produce a ground demonstrator that will be fully representative of robot functionalities and of the autonomy required for planetary exploration and therefore to develop the techniques that will allow the exploration and subsequent operation of planetary sites (Mars, the Moon). The following aspects relating to mobile robotics are developed :
- algorithms and systems for travel over natural and ill-known terrain,
- equipment required to autonomously determine the robot's position,
- equipment required for autonomous on-board environment perception,
- aspects related to on-board decision-making and the planning of operations associated with the operator station on Earth,
- handling and gripping tools.
The important perspectives that would open the ESA initiative of a Lunar scientific base to be situated in the South pole area as well as the difficulties in scheduling an ambitious exploration of Mars, have lead CNES to shift the current thrust of the work carried on within IARES to a lighter demonstrator, and to lead work and validation experiments aimed to anticipate and to support a first technical Lunar mission, such as LEDA or Euromoon, by the beginning of the 21st century.
The IARES demonstrator, whose locomotion platform and control have been tested on the GEROMS test site, will be equipped with a manipulation arm and an inertial unit in December 1997 and will start system-level tests mid-1998.
Future activities will then be oriented towards development of space qualified components and sub-systems.
An important international science-oriented research effort is conducted in Antarctica. France (IFTP) and Italy (ENEA) have programmed a major undertaking : the CONCORDIA project.
A scientific base, Dome C, is to be installed to carry on an ambitious scientific programme in the heart of Antarctica. Dome C is situated at a similar distance (approximately one thousand kilometers) of the Italian coastal base and the French one (Dumont d'Urville). Research groups in both countries have started studies related to robotics in two possible directions :
- Logistics uses : coastal base - Dome C traverses
- Scientific experiments : around Dome C ; teleoperated experimentations, automatic monitoring,...
For the French side, RISP, that was entrusted in 1996 with Robotics studies and now LAAS-CNRS that has taken over this task, have focused their interest mainly in science-oriented robotics requirements.
The work in progress has been two-folded. First, careful consideration of the foreseen scientific experiments has been carried out by the international experts involved in the project CONCORDIA. So far, among the most interesting aspects relevant to Robotics assistance, two appear to be highly interesting:
A mobile robot RISP-C, with an original robot architecture has been designed in cooperation with RCL, a Russian (St-Petersburg) company subsidiary of VNIITRANSMASH.
Current work is conducted in cooperation with the ENEA (Italy) Robotics group, that has been developing a robotized Kassborer for logistic uses. In particular, the research cooperation covers the subjects related to the functional architecture, the perception systems, and with a special emphasis, the instrumental payload.
CEA Commission de l'Energie Atomique
CREA Centre de Recherche et d'Etudes de l'Armement
EDF Electricité de France
IEF Institut d'Electronique Fondamentale
IFREMER Institut Français de Recherche et d'Exploration de la Mer
I3S Laboratoire d'Informatique Signaux des Systèmes de Sophia-Antipolis
ITTP Institut Français pour la Recherche et la Technologie Polaires
INRETS Institut National d'Etudes et de Recherches sur les Transports
INRIA Institut National de Recherche en Informatique et Automatique
INSERM Institut National de la Santé et de la Recherche Médicale
IRISA Institut de Recherches en Informatique et Systèmes Aléatoires
IRIT Institut de Recherches en Informatique de Toulouse
JSPS Japanese Society for the Promotion of Sciences
LAEI Laboratoire d'Automatique et d'Electronique Industrielle
LAAS Laboratoire d'Analyse et d'Architecture des Systèmes
LAG Laboratoire d'Automatique de Grenoble
LAI Laboratoire d'Automatique Industrielle
LAN Laboratoire d'Automatique de Nantes
LASMEA Laboratoire des Sciences et Matière pour l'Electronique et d'Automatique
LCPC Laboratoire Central des Ponts et Chaussées
LIAH Laboratoire d'Automatique et de Mécanique Industrielle et Humaines
LIFIA Laboratoire d'Informatique Fondamentale et d'Intelligence Artificielle
LIRMM Laboratoire d'Informatique, de Robotique et de Microélectronique de Montpellier
LRP Laboratoire de Robotique de Paris
ONERA Office National d'Etudes et de Recherches Aérospatiales
RSM Robotique Sous-Marine Méditerranée