European Commission logo
español español
CORDIS - Resultados de investigaciones de la UE

ACANTO: A CyberphysicAl social NeTwOrk using robot friends

Periodic Reporting for period 3 - ACANTO (ACANTO: A CyberphysicAl social NeTwOrk using robot friends)

Período documentado: 2017-05-01 hasta 2018-07-31

The overarching goal of ACANTO is to develop a combination of technologies to spur older adults into a regular and sustained level of physical activity. ACANTO's solution rests on two pillars:
1. a robotic friend (the FriWalk) to support the execution of everyday activities,
2. an intelligent system, called Cyberphysical Social Network (CPSN), that produces recommendation for social activities matching the profile of the users, as they result from the observation collected on the ground through the FriWalk.
The FriWalk is robotic walker embedding a large portfolio of solutions to sense the surrounding environment, plan the most comfortable and
safe routes to the user's destination, and, where required, gently guide the user in her navigation of the space. The robot constantly
observes its user noting all relevant changes in her physical and emotional state. The observations on the user state are consolidated into her personal profile. Furthermore, the FriWalk can be used in the context of clinical rehabilitation to diagnose the condition of the patient, to guide her in the execution of rehabilitation exercise and to follow the evolution of her clinical conditions.
The CPSN collects the profiles of the user and aggregates them into circles, recommending social activities to each circle. The objective of the CPSN is to stimulate the user
social life offsetting the effects of loneliness.
In the context outlined above, the concrete goals of the project are:
1. development of the FriWalk: from the sensing subsystems to the mechatronic components and to the user interface, 2. development of the recommendation system and of all the software components animating the CPSN,
3. development of an external infrastructure (cloud and environment sensing) that supports both the conception and the execution of social activities,4. large involvement of users and stakeholder in the development of the different components and in the definition of realistic exploitation routes.
During the execution of the project, the ACANTO consortium has:
1. established the project's management structure and defined data management and quality plans.
2. Completed the collection of requirements and the study on Privacy issues. Such requirements are related to the application of our both in a "social" context (e.g. a Museum) and
in a clinical environment.
3. Defined the basic procedures for Ethics management and their overarching principles. Based on these principles, we have produced a number of Ethic Clearance applications required by the different user studies.
4. Integrated a first group of functionalities into a first version of the prototype, which has been used to refine the requirements.
5. identified and developed a second group of functionalities, which have been integrated in the final version of the prototype of the walker.
6. Studied the state of the art on dynamic models that describe user group motion in competitive and cooperative situations and sketched the ACANTO solution, which have been implemented in the reactive planner of the walker.
7. developed the software models for user profiles, circles and activities.
8. completed the development of CPSN and recommendation systems based on the software models and on a cloud infrastructure.
9. Developed and refined the user interfaces that will be used both for the clinical applications of the FriWalk, and for the public applications of FriWalk and CPSN.
10. Set up the framework for clinical validation (e.g. certification of the walker and Ethics management procedures).
11. Developed a clinical pilot study, in which the efficacy of the FriWalk has been evaluated on 42 patients in SPain and tested on 10 patients in Italy
12. Validated the use of the FriWalk, and of the haptic interfaces as a tool for navigation assistance on a large number of older adults (68 in Trento and 30 in New Castle)
13. Tested the recommendation system on a sample of senior users (around 20) and on a large number of members of the public using microwork platform
14. Executed a longitudinal study on 35 users to validate the general assumptions underlying the ACANTO project
15. Produced two complete business plans that define possible exploitation routes
16. Executed a wide range of communication activities toward the scientific community, the stakeholders and the members of the public. The scientific papers published in reputable journals are 18 (+5 subimssions), while 52 papers have been published in top conferences
17. Defined policies and executed concrete actions for IPR management (5 patents request have been filed and two more are in prepatration).
During the course of the project, we have:
1. developed an innovative framework that mapped privacy concerns to the "Johari window" technique to show how older adult privacy preferences relate to different categories of data about themselves. We pointed out that older adults have privacy concerns about data relating to themselves that even they do not want to be aware of (e.g. data about their health decline).
2. identified a number of “rules” to drive a proper interaction between the researcher and the user in the extraction of requirements for assistive robotic systems, which has been at the heart of several journal publications.
3. developed a model to predict the motion of human agents present in the environment.
4. developed localisation algorithms that fuse information from different sources (encoders, IMU, markers in the environment, SLAMMOT).
5. proposed several techniques for gait analysis, human body tracking, Slammot and emotion detection, which have brought significant innovation in their respective fields (as witnessed by the several papers).
6. used the the perception algorithms along with our robotic platform for the clinical application (one of the most important point of innovation of the project).
7. developed the CPSN and the Recommendation algorithms, which are one of the first examples of a social network that: 1. explicitly target the needs of the senior population, 2. Are interconnected with the user’s clinical profile and needs, 3. Account for the tastes and the requirement of the person emerged through physical observations.
8. developed an Activity Planner (WP5) that plans a sequence of actions to implement an activity. In the same context, we developed a Reactive Planner to compute an optimal motion plan for a robot moving in an environment crowded by humans.
9. developed several guidance solutions that attracted the attention of the robotic community, were extensively tested with users and put ACANTO’s team at the fore front of innovation on assistive robotics. In addition, a large effort has been allocated to the development of guidance and signalling devices based on haptic feedback.
10. produced algorithms and tools for probabilistic guarantees for soft real—time systems.
These innovations have brought to more than 70 scientific papers and 5 patent submissions (with two more in preparation). We produced a first evidence of the clinical benefits of using the FriWalk, in: 1. improving the condition of the patient, 2. reducing the effect of traumatic events and the probability of their re-occurrence, 2. contacted a large number of stakeholders and policy-makers who encouraged us in seeking concrete exploitation routes for the ACANTO technology.
ACANTO new logo
ACANTO new brochure