The objective of Loco3D is to build a methodology to execute a contact sequence, computed with an efficient motion planner, with a powerful humanoid robot inside a complex environment subject to dynamic changes. Planning and adapting complex locomotion patterns is a key problem that prevents from releasing humanoid robots and other legged manipulators in non-conventional factories, such as the assembly workshop of the Airbus A350 and A380 aircrafts in Toulouse, and beyond, the implementation of rehabilitation exoskeletons (e.g. for paraplegic) and the use service humanoid robots in offices, hospital and home environments. In order to achieve this scientific objective, we divided the project in two stages, each one having its own scientific targets and outcomes. The first stage relies on mature scientific methods developed by our team, and serves as a basis for an original approach, in rupture with existing works, in order to tackle the locomotion problem in dynamic environments in a second stage. It also aims at bringing machine learning within the research topics of our team. In the first stage, we will consider the nominal problem of locomotion in a complex static environment. The difficulty of this problem was emphasized by the recent DARPA Robotics Challenge and the spectacular falls of most of the biped robots involved. Based on some recent preliminary developments (on top of which the first real-time contact planner and the first real-time 3D locomotion pattern generator), we believe that our team is able to propose the premiere complete solution to this problem. The first part of the project will lead to the development of a first demonstrator that will be used as a springboard for the second phase of the project. This first phase heavily relies on careful heuristics and domain-specific developments to handle locomotion in static environment. It might be possible to extend it to dynamic locomotion in uncertain environment, however asking again for years of development to find the proper heuristics and dedicated motion models able to handle these new hypotheses. Rather than doing the same work a second time in another context, we propose to study the issue of automatically generalizing the said heuristics in any motion context, by relying on the automatic construction of a robot motion experience. Using the current know-how in robotics, it is relatively easy to automatically generate an efficient movement for any complex dynamics if the computation time is not considered. The question becomes much difficult when considering the hard time constraints imposed by the control frequency on a real hardware. A logical solution is to rely on pre-computation and storage of possible control values. However, the typical size of a naive storage is not acceptable. We propose in the second part of the project to transform the problem of motion generation in robotics into a problem of big data, by studying how to reduce the size of such a pre-computation dataset. The stake in data reduction is to exhibit a sparse structure in the dataset. The mathematical structure underlying the control function is explained at best by the Hamilton-Jacobi-Bellman equation. We propose to exhibit its sparsity by writing the reduction problem as an extended inverse optimal control problem. The project is built on the applicative scenario of legged locomotion in complex dynamic environments. However, the general methodology is generic and would apply to the control of other complex dynamics. We will demonstrate the interest of using such an automatically-built robot experience by solving the problem of locomotion in dynamic environment, through the implementation of a complete demonstrator on the new humanoid robot of the laboratory. We will also demonstrate that it is generic by applying the same method to execute complex maneuvers with aerial manipulators in simulation.

We recently developed, patented, and successfully transferred a technology for DNA purification, enrichment, and separation. This technology coined µLAS involves an electric field and a counter hydrodynamic flow in viscoelastic liquids, in which transverse forces oriented toward the walls occur. These forces increase with DNA molecular weight (MW) and hence induce a progressive reduction in DNA migration speed that triggers size separation in microfluidic channels as well as in capillaries. Therefore with conventional microfluidic control systems of pressure and electric field, the transport of DNA can be finely controlled. More specifically using commercial Capillary Electrophroresis, this technology allows us to perform DNA separation in the 0.1-5 kbp with unrivalled sensitivity of 20 pg/mL with an operation time of ~10 minutes. In this proposal we aim to bring this technology one step ahead and perform the operations of separation, enrichment and purification for virtually every DNA molecular weight. We target academic and industrial needs in third generation sequencing, bacteriology, epidemiology, and cancer diagnostics. For this we will optimize the separation mechanism according to rational rules determined by specific physics models of flows in microchannels. More specifically, we intend to perform original experimental and theoretical researches on visco-elastic lift forces using different families of polymer solutions (WP2). Our goal is the development of a predictive platform to reach µLAS optimal performances. We will then confirm or invalidate the predictions of our platform by running separation experiments with DNA molecules of increasing molecular weight (WP3), and obtain optimal separation performances for the different DNA size ranges targeted in this project. We then wish to investigate whether enhanced separation performances for high molecular weight molecules can be reached with temporal modulations of the electric field (WP4). Note that WP3 and WP4 are mutually reinforcing: both aim to gradually improve the features of µLAS for DNA separation. The final task of this project (WP5) is devoted to the specification of a prototype for strain typing and/or quality control of bioprocesses, such as third generation DNA sequencing. Our project will be accomplished by LAAS and LRP, which are two laboratories expert in DNA separation in microfluidic systems and complex fluids hydrodynamics, respectively. Developments will be transferred to an SME company Picometrics for industrial prototyping.

In a foreseeable future, humans and robots are set to act together and even to collaborate. Service, assistant or teammate robots are actively considered as possible applications. Collaborating shall then not be tied to one single, specific interaction, since all the agents concerned, both the human AND the robot, will then have to build and to maintain, co- constructively, and as long as necessary, the collaborative process and relationship that come along with the task, thus allowing its joint execution. Such a process must be flexible, adaptable enough, so any of the agents, human or (even) robots, could make mistakes, fail locally, speed up or slow down, change their "mind", modify their decisions, etc., along the way, without undermining the interaction. In order to make it possible, all those elements have to be recognized just for what they are. Considerable work is currently ongoing, addressing those points and contributing to make this interaction easier and smoother; and we take part of it. We envision this work as a first step on which the project JointAction4HRI will be able to build a solid and comprehensive theoretical framework to help human-robot interaction design. This is the way we are aiming to orient our research. We strongly believe that a human agent should be able to engage into an interaction with a robot in a natural way, with some social references she is familiar with, that she knows how to use and to deal with, and that are those she would use in a usual interaction with another human. Systematically identifying these references, understanding how they work together, adapting and integrating these social skills to the robot are the root constituents of the project JointAction4HRI. Far beyond the "simplistic" anthropomorphic appearance or the "naive" human-mimicry that we could implement to the machine, what matters to us here are the key components and mechanisms underlying joint action, their informational content (relatively to the common ground in terms of general knowledge, acquired previously and during the task itself), and how they are linked together, temporarily, hierarchically and meaningfully. Our ambition is to contribute, at the most fundamental level, to the identification and characterization of the ingredients required for the successful realization of a collaborative task. The JointAction4HRI project will address this goal through three objectives: (1) Identifying the elements (and their interaction) that allow the establishment and maintenance of mutual understanding between a human and a robot in a collaborative task (2) Modeling the notion of commitment that structures the relationship between two agents during interaction (3) Elaborate and test robotics architecture design for collaboration with humans. It proposes an original consortium composed of researchers from a philosophy laboratory (Institut Jean Nicod), a psychology laboratory (CLLE), and a robotics laboratory (LAAS). Each has contributed in their respective discipline and for several years to research in the joint action domain. By bringing their complementary expertise together, JointAction4HRI will foster the development of an integrated theoretical framework. Expected benefits are obvious from the perspective of service or entertainment robotics development. But we can envision broader benefits. Thus, we expect that joint attention modeling will give us cues to help the evaluation and support of children with developmental disorders (e.g. autistic children). Confronting the roles of commitments in human-human interaction and in human-robot interaction should also shed new light on the current philosophical debate about commitments in joint action, helping to identify more precisely their essential properties as well as the nature of the social norms and emotions that ensure their efficiency.

The main objective of MIMIC-SEL is to create the first-ever electrically-pumped Vertical-Cavity Surface Emitting Laser (VCSEL) technology emitting at wavelengths longer than 3µm and operating in CW mode at temperature higher than 300K. The availability of electrically pumped VCSELs at wavelength longer than 3 µm operating in continuous wave above room temperature is considered as a breakthrough for laser-based optical sensing applications. Indeed, these devices have a low cost potential, low power consumption and overall emits single frequency with mode-hop free tunability. These ideal and unmatched properties will enable widespread utilization of photonic sensor networks. This will contribute to the well being and health of the population by stimulating air pollution monitoring, detecting leaks and preventing fire as other possible application. Electrically-pumped mid-IR VCSELs rely on GaSb-based materials. But entering in the 3-5 µm wavelength range requires new approaches for both active zone design and device processing. Thus, the structure will be based on an interband cascade type-II active zone, a metamorphic oxide-based lateral confinement scheme and a hybrid-mirror technology with target performances suited for spectroscopy applications. The starting point of this project is the complementary knowhow of the consortium with the expertise of IES on the growth and study of GaSb-based VCSELs, the capability of LAAS for VCSELs processing, in particular regarding the controlled oxidation of Al(Ga)As layers, and of FOTON for their expertise in dielectric materials dedicated to Bragg mirror. The project is organized into 4 tasks, one for the management and 3 for technical development. The first part (Task 2 and 3) of the MIMIC-SEL project will develop every element constituting the VCSEL structure, i.e. the active region and the Bragg mirrors for operation beyond 3 µm. This part will give us the information on the key parameters such as gain, reflectivity, optical losses, and thermal properties to properly design the VCSELs. The last part (Task 4) will be related to the realization, characterization of VCSELs and application to gas sensing.

Several diseases spread between persons via the interaction of our bare hands with the outside world. Indeed, a large number of pathogens can survive on inert objects and surfaces and be later transferred onto a person hand who will then auto-contaminate herself. High morbidity or high mortality disease transmission through the contact of our hand with contaminated fomites is responsible each year of a large amount of hospitalizations, with a high associated societal and economical cost. At the present time, owing to a lack of appropriate technology, it is impossible to track down the path of a pathogen being deposited onto a fomite and transferred onto our hands with portable, automated equipment. Indeed, portable miniaturized biological sensors available are only able to detect pathogen at relatively high concentrations. Typically, the measurable pathogen concentrations of current portable equipment correspond to the concentration levels one would find inside the body fluids of an already sick person. So there is currently no fast and portable method to detect pathogen presence on field before they have contaminated and multiplied into a person, which limits disease and biological risk surveillance to “detect-to-treat” approaches. This proposal aims at demonstrating the proof-of-concept of an innovative biological detection and identification technology that would allow the sensing of few pathogen units inside of middle-size liquid volume (typically the volume needed to rinse one’s hands). The short term goal is to combine a pathogen pre-concentrator system with an array of MEMS liquid biological sensors to demonstrate sensing of low viral particles concentrations within a short response time. Neither of these approaches has ever been implemented before to the consortium’s best knowledge. As a matter of fact, the challenge is significant to provide a MEMS sensor that at the same time exhibits good mass sensitivity, a large capture area and that has been properly functionalized with high specificity antibodies to avoid at maximum the occurrence of false-positives. Challenges are also numerous to implement a fluidic tool enabling pre-concentration of viral particles at low concentrations within large volumes and record timing. To overcome these challenges with success, a strong consortium has been built between LETI, LI2D and LASS. This team has combined expertise in micro/nano fluidic systems, MEMS/NEMS based biological sensors, biological functionalization protocols and in design and production of highly sensitive and specific monoclonal antibodies towards specific pathogen viral strains. The long term goal of the proposed research is to provide an upper TRL technological unit enabling fast, ultra-sensitive and fully reliable detection method for pathogens. Such a key enabling technology could be applied, as an example to detect pathogen sampled directly from persons’ hands. Being able to detect pathogen on the hand of a still uncontaminated person would unlock many potential applications and will open a new paradigm in epidemiological surveillance.