Progetti

Natural Intelligence

Natural Intelligence

Il riscaldamento globale e l'inquinamento stanno minacciando la sopravvivenza di un milione degli otto milioni di specie nelle foreste e negli oceani del pianeta. La risposta dell'UE a queste sfide è l'insieme di politiche innovative contenute nell'European Green Deal. Tra queste politiche un ruolo di primo piano è dato al ripristino e alla conservazione degli ecosistemi, aumentando la copertura delle aree terrestri e marine protette ricche di biodiversità sulla base della rete Natura 2000. Estendendosi sul 18% della superficie terrestre dell'UE e quasi il 6% del suo territorio marino, Natura 2000 è la più grande rete coordinata di aree protette nel mondo. La copertura terrestre e marina sarà aumentata fino al 30% entro il 2030. Oggi gli operatori umani sono l'unica opzione per eseguire il monitoraggio ambientale di un'area così vasta. Questo non solo per la loro specifica competenza e conoscenza nella classificazione delle piante e degli habitat, ma anche per la loro intelligenza fisica che permette loro di muoversi per ore in ambienti selvaggi non strutturati come dune, foreste e montagne. L'alternativa artificiale è la robotica, che ha fatto enormi progressi negli ultimi anni, tuttavia i robot difficilmente lasciano i laboratori e le fabbriche perché non sono robusti ed efficienti per sopravvivere nel mondo reale. I robot volanti non hanno l'autonomia energetica necessaria, mentre i robot di terra non sono:

i) intelligenti per percepire, interpretare e interagire autonomamente con terreni altamente irregolari e scivolosi;

ii) fisicamente robusti per gestire contatti e impatti imprevisti.

Il progetto Natural Intelligence (NI) mira a servire il Green Deal europeo attraverso il monitoraggio degli habitat naturali di N2000N con robot in grado di muoversi efficacemente in dune, praterie, foreste e terreni alpini. I robot NI saranno potenziati dalla Natural Intelligence, che emerge dall'interazione di ambiente, corpo e mente, facendo leva sulla fusione di cognizione artificiale e corpi articolati di soft-robotica.

ReconCycle – Self-reconfiguration of a robotic workcell for the recycling of electronic waste

ReconCycle – Self-reconfiguration of a robotic workcell for the recycling of electronic waste

Attualmente i robot industriali eseguono compiti rigidamente programmati in ambienti altamente limitati. Ogni cambiamento nel prodotto o nel processo richiede una costosa ristrutturazione di hardware e software.

ReconCycle affronterà questi problemi introducendo il concetto di auto-riconfigurazione robotica nel dominio largamente non vincolato del riciclaggio dei rifiuti elettronici, che è ancora dominato dal lavoro manuale. L'automazione in questo settore può trarre vantaggio dal fatto che devono essere trattati lotti molto grandi dello stesso tipo di dispositivo, ma con alcune differenze di modello e che mostrano diversi stati di danno. Per essere in grado di trattare ognuno di questi modelli individuali, il sistema robotico richiede un adattamento flessibile.

Così, l'obiettivo scientifico di ReconCycle è di introdurre in questo settore hardware e software auto-riconfigurabile basato su una cella robotica riconfigurabile sviluppata in un progetto precedente. È prevista una procedura in due fasi: Quando si passa da un tipo di dispositivo ad un altro, la riconfigurazione sarà eseguita in una modalità interattiva, dove l'ingegnere dell'applicazione sarà in grado di fornire input. Ma, quando si passa da un modello di dispositivo a un altro all'interno di un dato tipo di dispositivo, la cella eseguirà la riconfigurazione da sola attraverso una combinazione di approcci di apprendimento sensorimotorio e altre tecniche di IA. Questo costituisce il principale contributo scientifico di ReconCycle e mira a far progredire la nostra comprensione dei processi di percezione-azione robotica in ambienti industriali non vincolati. Useremo robot morbidi ed end-effettori altamente conformi, permettendo agli umani di operare insieme alle macchine per completare qualsiasi passo mancante. Questo riduce la complessità dell'automazione e porta questo progetto in un regime fattibile fino al TRL 6. Il riciclaggio dei rifiuti elettronici è un settore economico e ambientale importante e in forte crescita.

L'obiettivo industriale di ReconCycle è di introdurre qui un livello molto più elevato di automazione con un impatto potenzialmente alto a lungo termine sull'industria e sulla società.

qbrobotics Progetto “Espansione internazionale e consolidamento della presenza di QBROBOTICS nel settore della robotica collaborativa negli Stati Uniti e in Cina” finanziato dal POR FESR Toscana 2014-2020

qbrobotics Progetto “Espansione internazionale e consolidamento della presenza di QBROBOTICS nel settore della robotica collaborativa negli Stati Uniti e in Cina” finanziato dal POR FESR Toscana 2014-2020

qbrobotics ha ottenuto il finanziamento di alcune iniziative di espansione internazionale grazie al Bando Internazionalizzazione promosso dalla Regione Toscana nell'ambito del Programma operativo regionale (Por) Fondo europeo di sviluppo regionale (Fesr) 2014-2020.

Il progetto denominato “ESPANSIONE INTERNAZIONALE E CONSOLIDAMENTO DELLA PRESENZA DI QBROBOTICS NEL SETTORE DELLA ROBOTICA COLLABORATIVA NEGLI STATI UNITI E IN CINA” ha supportato le iniziative promosse dall'azienda nei due Paesi target (Stati Uniti e Cina).

In particolare, oggetto del finanziamento è stata un'attività di comunicazione e la partecipazione a fiere di rilevanza internazionale che hanno avuto lo scopo di sostenere l'espansione internazionale di qbrobotics nei due mercati di riferimento. I risultati attesi sono stati ampiamente confermati e raggiunti grazie all'implementazione di una rete distributiva strutturata e ad un conseguente incremento del fatturato generato.

EUROBENCH project – DYSTURBANCE: Dynamic and static pusher to benchmark balance

EUROBENCH project – DYSTURBANCE: Dynamic and static pusher to benchmark balance

An old adage in the field of robotics says that “the best measure to assess the robustness of a robot system is how far you trust to be when the system is functioning”. This reflects both the fact that one of the most important aspect that distinguishes a good robotic system from a mediocre one, is its robustness to uncertainties and unexpected events, and the fact that measuring this difference is still mostly done with rules of the thumb. Both of these considerations are still very true for legged locomotion system. Important aspects that characterizes a legged robot system are the robustness of its static balance and of its dynamic walking gait. The DYSTURBANCE project proposes to realize a perturbation system to be used for the systematic characterization of walking robots and locomotion systems. In particular, the system we plan to realize will enable reproducible issuing of both dynamic and static perturbations forces, that would enable the systematic study of gait and balance resilience and robustness. The proposed system draws inspiration from classical Resilience testing machines used in standardized testing of material samples (as the Charpy test rig) so to realize, in the simplest and easily reproducible way, a system capable of exerting a given desired force or impulse perturbation. The system will integrate position, and force sensors, so to be able to fully characterize the disturbance that we are applying to the walking system. Moreover, it will integrate a motorized shaft, so to be able to control the exertion of forces both in the static and dynamic cases. Finally, the system will be implemented in the form of a moveable portal for easy integration with other experimental rigs, as treadmills, sensorized platforms and other system.

SoftHandler – Commercial feasibility of an integrated soft robotic system for industrial handling

SoftHandler – Commercial feasibility of an integrated soft robotic system for industrial handling

"Picking and placing objects that are orderly presented to a robot is a task that can be easily performed by current robots. However, grasping and manipulating objects randomly placed in an unstructured environment remains one of the hardest challenges for robots. Examples of complex picking industrial applications include (i) bin picking for part feeding, (ii) raw food handling and (iii) waste sorting for recycling. Within the ""SoftHandler"" ERC Proof of Concept project, we propose to develop new industrial-grade systems, comprised of end-effectors based on the ""SoftHands"" technology, integrated with a novel manipulator, for automated picking and placing of objects of heterogeneous dimension, shape, weight, position and strength. The Proof of Concept project will be focused on building a commercially viable prototype, and demonstrating it in real-world scenarios.

THING – subTerranean Haptic INvestiGator

THING – subTerranean Haptic INvestiGator

THING will advance the perceptual capabilities of highly mobile legged platforms through haptic perception and active exploration. In this light, THING will deliver:

1) Novel foot designs for enhanced tactile perception and locomotion,

2) Improved perceptual capability, enriching existing modalities (lidar, vision) with haptic information,

3) Heightened physical sense of the environment, including friction, ground stability (difficult through vision alone), and

4) Enhanced mobility through improved perception, prediction, and control.

SoftHand Pro-H – A Soft Synergy-based Hand Prosthesis with Hybrid Control

SoftHand Pro-H – A Soft Synergy-based Hand Prosthesis with Hybrid Control

In the project “SoftHands” we have achieved not only a more thorough understanding of the organization and control of hands, but also a principled approach to taming the complexity of hand design. The original concept of “soft synergies” has underpinned the realization of radically new artificial hands: “SoftHands” have been demonstrated to be more adaptive and capable than most artificial hands, yet are simpler to control and more robust. One application of these ideas and technologies which stands out for potential impact and social relevance, although not originally foreseen in the ERC AG plan, is the realization of upper limb pros-theses. The objective to realize a prosthetic hand that is anthropomorphic, aesthetically pleasing, and enables an amputee to perform most activities of daily living as well as advanced prostheses, while being robust, intuitive, and economic as basic body-powered split-hook prostheses requires a much longer and larger re-search and development effort than an ERC POC can support. In this proposal we study the feasibility of applying the SoftHand technology to address one particular, but very important, objective, i.e. work-oriented prostheses. The specific requirements of these applications are high grip power, grasp versatility, resilience, resistance to water, dust, and temperature, durability, power autonomy and low cost – while factors such as aesthetics or silent operation are less dominant. Of particular relevance is the control interface with the patient. Virtually all work-oriented prostheses are operated via a body-powered cable, which is very intuitive to use and does not need batteries, motors, and sensors. On the other hand, advanced multi-fingered prostheses have sophisticated myoelectric control affording versatility and dexterity. In this project, we will engineer and experiment a novel hybrid control for a SoftHand prosthe-sis, whereby a traditional cable harness commands the advanced mechatronic system of the SoftHand.

Daedalus – Distributed control and simulAtion platform to support an Ecosystem of DigitAL aUtomation developerS

Daedalus – Distributed control and simulAtion platform to support an Ecosystem of DigitAL aUtomation developerS

Daedalus is conceived to enable the full exploitation of the CPS concept of virtualized intelligence, through the adoption of a completely distributed automation platform based on IEC-61499 standard, fostering the creation of a Digital Ecosystem that could go beyond the current limits of manufacturing control systems and propose an ever-growing market of innovative solutions for the design, engineering, production and maintenance of plants’ automation. The following objectives will be reached: -)Ease the conception, development and distribution of intelligence into CPS for real-time execution of orchestrated manufacturing tasks; -)Foster interoperability of CPS from different vendors at orchestration-level (= “between CPS”); -)Simplify the design, implementation and integration of optimal coordinating control intelligence of CPS; -)Enable near-real-time co-simulation of manufacturing systems as a fully integrated “service” of a CPS; -)Create a Digital Marketplace to simplify the matchmaking between offer and demand within the Ecosystem; -)Conceive a multi-sided business model for the Automation Ecosystem and the corresponding business plans for its Complementors; -)Foster the widespread acceptance of the Ecosystem platform to guarantee success and impact of Daedalus multi-sided market. The project approach is based on 3 technological pillars, one platform pillar and a final ecosystem pillar: 1)IEC61499-based reconfigurable automation platform for distributed orchestration of interoperable CPS; 2)“Simulation-as-a-service” for integrated near-real-time co-simulation of CPS behavioural models; 3)Advanced SDKs for simplified design of hierarchically distributed optimal control applications; 4)Digital Marketplace for the creation of an interdependent ecosystem of automation solutions providers; 5)Proof-of-concept showcases to accelerate the involvement of “complementors” and the maturation of the Ecosystem.
SoftPro – Synergy-based Open-source Foundations and Technologies for Prosthetics and RehabilitatiOn

SoftPro – Synergy-based Open-source Foundations and Technologies for Prosthetics and RehabilitatiOn

Although much has been done for developing technologies to bear upon problems of individuals with sensorimotor impairments, the impact of robotic aids on people with real needs in the real world is still very limited. Our main goal is to increase the cumulative benefits of assistive robotic technologies to society by enhancing their effectiveness AND the number of beneficiaries. The challenge is to increase both multipliers in the “performance times accessibility” product, subverting the traditional situation where one factor can only be increased at the expense of the other. We believe this is possible by investigating how the artificial can physically interact and effectively “talk to” the natural. Understanding such a “language” is crucial not only to improve performance of rehab technology, but also to tackle the most difficult problem of making it “simple enough” to be effective and accessible. We possess good clues about such a language, whose words we believe are sensorimotor synergies, and have the scientific competence to further its understanding and the technological prowess to translate it into a new generation of robotic assistive devices.
We know that a central ingredient for the applicability of synergy-based models to physical human-machine interaction is impedance adaptability, i.e. soft robotics technologies. We will develop soft synergy-based robotics technologies to produce new prostheses, exoskeletons, and assistive devices for upper limb rehabilitation.
Building on solid methodological bases, this project will have a significant social impact in promoting advanced robot prosthetic and assistive technology, while introducing disruptively new, admittedly risky, but potentially high-impact ideas and paradigms, such as the proposed pioneering work on supernumerary limbs for assistance and rehabilitation to motor impairments of the upper limb.

SAPHARI – Safe and Autonomous Physical Human-Aware Robot Interaction

SAPHARI – Safe and Autonomous Physical Human-Aware Robot Interaction

Cognitive Systems and Robotics Recent progress in physical Human-Robot Interaction showed that active and safe workspace sharing becomes possible in principle. SAPHARI will perform a fundamental paradigm shift in robot development in the sense that the human will be the centre of the entire design. The project will enable robots to track, understand and predict human motions in a weakly structured dynamic environment in real-time. The project will focus on industrial use cases that require contacts and force exchange in human-robot co-work, as well as on professional service scenarios in hospitals, in which a medical staff and an assisting robot interact closely during daily work. Results of this project will impact all applications where interactive robots can assist humans and release them from dangerous or routine tasks.