Superlerarning

High Energy Efficiency Biped Robot controlled by the Human Brain for people with ALS disease*

P. Fedele1, P. Federighi2, R. Molfino3, G. G. Muscolo3,4,a, C. T. Recchiuto5, A. Rufa2b

1Liquidweb s.r.l., Via XXIV Maggio, 21, Siena, Italy. 2Eye-tracking & Visual Application Lab Department of Medicine, Surgery and Neurosciences University of Siena, Siena,

Italy. 3PMAR Lab., DIME-MEC, Applied Mechanics and Machine Design, Scuola Politecnica, University of Genova, Via

all’Opera Pia, 15, Genova, Italy. 4Creative and Visionary Design Laboratory, Humanot s.r.l., Via Amedeo Modigliani, 7, 59100, Prato, Italy.

5Electro-Informatic Laboratory, Humanot s.r.l., Via Amedeo Modigliani, 7, 59100, Prato, Italy.

bAuthors listed in alphabetical order, equal contribution.

Abstract—This paper analyses the adoption of novel robotic technologies to help people suffering from amyotrophic lateral sclerosis (ALS). The work starts with the analysis of problems that mostly affect these patients in order to understand where science and technology can be used to help them. In particular brain-computer interfaces and their implementation with a high energy efficiency humanoid robot for domestic assistance will be taken into account. The features that this kind of technology must have in order to satisfy the required need will be discussed, with some preliminary tests implemented in existing platforms.

Keywords-ALS, Humanoid Robots, BCIs, Human-Machine Interface, Energy Efficiency, Cognitive Rehabilitation, Biomechanics.

I. INTRODUCTION The rapid development of robotic technologies in the last

thirty years has created a tight connection between the health care sector and robotic technologies. Advanced technologies have been applied to medical fields such as surgery, diagnosis, rehabilitation, prosthetics, and assistance to disabled and elderly people and the implementation of robotics artifacts in these sectors is a growing trend.

The use of robotic surgery systems in medical procedures dates back to the 1980s, aiming at reducing the invasiveness of interventions and ensuring a high level of accuracy [1]. In this context, robots can be designed in order to provide support to surgeons, performing tasks that are usually done by human assistants [2] or may be directly involved in surgical interventions, manipulating instruments and performing crucial stages of the process. Robots that are directly involved in surgical procedures can be designed to perform a specific intervention (e.g. PROBOT, specialized

in prostate interventions [3]), or can be “general-purpose”, with robotic arms equipped with various instruments (scissors, endoscopes,…) [4] and controlled by a surgeon by means of a tele-operated robotic system (e.g. Zeus, da Vinci [5] and CyberKnife [6]). The last frontier of robotic surgery is nanorobotics, which include nanodevices that can operate at molecular levels to, for example, reconstruct damaged tissues or identify and destroy cancerous cells [7].

Invasive and non-invasive robotics also play a fundamental role in diagnosis. Some examples include ecographic robotics systems [8] or robotic endoscopes included in the ZEUS and da Vinci systems [5]. The most modern approach is focusing on the development of miniature diagnostic robots endowed with locomotion abilities, including small energy sources and wireless communication devices [9].

Robotics is widely used also in the prosthetics field. Thanks to myoelectric signals, robotics arm or legs can be finely controlled. Examples are the i-Limb hand of Edinburgh or the European project Smarthand [10,11].

Another health-care sector where robotics is assuming a leading role is rehabilitation. Rehabilitation robotics focuses on machines that can be used to help people to recover from severe physical trauma. One quite recent tool for physical therapy in this field is the robotic exoskeleton. The most promising one being used is the LOPES (LOwer-extremity Powered Exo Skeleton) [12]. The LOPES system is intended to help people to regain lost motor control, such as after a serious stroke. On the other side, upper limb rehabilitation robots which are typically constituted by robotic arms linked to the patient’s upper limbs. One example of this kind of structures is the MANUS system of the MIT [13].

But robots can be used also in another typology of rehabilitation, such as cognitive rehabilitation therapies. Cognitive rehabilitation therapies usually use animaloid robots: for example, the robot PARO, which is shaped like a baby harp seal, is an advanced interactive robot developed by AIST that has the ability of reduce patient stress and stimulate the interaction between patients and caregivers

* This joint research was partly supported by the BrainHuRo Italian Project (Tuscany POR-CREO Fesr 2007-2013 – Le ali alle tue idee) and partly supported by University of Siena, Liquidweb s.r.l. and Humanot s.r.l. as active partners in the BrainHuRo project.

a Corresponding author: Giovanni Gerardo Muscolo ([email protected]) ([email protected])

17th IEEE Mediterranean Electrotechnical Conference, Beirut, Lebanon, 13-16 April 2014.

978-1-4799-2337-3/14/$31.00 ©2014 IEEE 386

[14]. PARO has five kinds of sensors in order to perceive people and the environment, and it can interact with people by moving its head and legs, making sound and showing a preferred behavior. For this purpose, robot can have also a human shape: the F.A.C.E. system, developed at the University of Pisa, is a human android, provided with a simple communication interface focused on the reproduction of facial expressions and meant to meet the patients affective needs, building a close relationship and stimulating the interaction capabilities in its users [15].

Humanoid robots can also be extremely useful when applied in therapies with children suffering from autism spectrum disorders (ASD) [16]. People with autism generally feel comfortable in predictable environments and enjoy interacting with computers and robots [17]. Many projects that see a strict connection between children suffering from ASD and humanoid robotic artifacts have been carried out and are currently being developed: the AuRoRa project used child-sized humanoid robot (Kaspar, Robota) to engage children in coordinated and synchronized interactions with the environment thus helping them to develop and increase their communication and social interaction skills [18]. The humanoid robot NAO was also used for these studies, achieving a 30 percent increase in social interactions and verbal communication when the robot was in the same room with children [19].

Based on all these good results of the connection between health-care and robotics, the authors believe that people suffering from Amyotrophic Lateral Sclerosis can also benefit from the assistance and the help of humanoid robotics technologies. Amyotrophic Lateral Sclerosis (ALS) is a debilitating neurological disease characterized by progressive weakness and muscle atrophy.

No cure has yet been found for ALS. The principal treatments for ALS are designed to relieve symptoms and improve the quality of life of patients. Clinicians often expect that ALS patients will experience depression and anxiety during the course of the disease [20]. In these cases patients are usually treated with drugs having an antidepressant or anxiolytic effect with inconstant and often partial efficacy. A cognitive and behavioral approach can be applied also in these cases, encouraging engagement in activities that are still practicable and pleasant. Advanced robotics can be applied also for these patients with a dual objective: assistive to improve the quality of life and behavioral motivating the use of the residual abilities. In particular, we suggest the implementation of Brain-Computer Interface technologies and humanoid robotics as personal assistance.

Brain-Computer Interfaces (BCIs) enable to connect computers and robotic devices with the human brain [21]. The sensors to detect the activity of the brain can be invasive (inserted deep inside brain tissues), partially invasive (under the scalp but externally to the cerebral cortex) or non-invasive (resting on the scalp). Non-invasive configurations are easier to wear, does not require surgery and cost less but generally produce a weaker signal quality. Another fundamental issue related to BCI concerns the choice of the neural properties to be recorded.

Some tests were already conducted with patients: they were able to control domestic robotic devices and drive

movements of a robotic arm [22]. In a different research project, BCI technologies were used by the patients to control a wheelchair [23]. The results showed the feasibility of continuously controlling complex robotic devices using non-invasive BCI. These technologies can therefore be applied to assistive and humanoid robots.

Some artificial intelligence systems have already been implemented in order to assist disabled and elderly people, but they were rarely inspired by the human anatomy. Humanoid robotics took its first step in Japan in 1973 with the WABOT-project. In Europe, the research related to humanoid robots is in a fast expansion; the German ARMAR III [24], a robot with head, torso and hands mounted on a motorized platform, is an example of this increasing interest of the European research centers. The EU founded RobotCub [25] research project is another great example; it allowed the design and development of a humanoid robot with characteristics similar to a child.

Many of these studies were synthesized in the realization of the humanoid platform WABIAN 2-R, and then in its Italian version SABIAN (Sant’Anna BIped humANoid) [26, 27, 28].

Based on these premises, the aim of this work is to analyze the problems and the needs related to people suffering from ALS and describe how BCI technologies and humanoid robotics can help to solve both practical and psychological issues. In particular, the collaboration among neuroscientists, engineers and computer scientists lead to the realization of a project called BRAINHURO (start: December 2012 - end: December 2014), in which techniques of a BCI system, are integrated with a humanoid robot in order to assist ALS patients and provide them with a user friendly instrument that makes their daily living activities quite autonomous and facilitates communication with other people. This approach makes easier the interaction of patients with people even in very advanced conditions such as the looked in syndrome.

The paper is structured as follows: Section II describes in detail the most important issues not yet solved related to locked-in patients, and their most urgent needs. Section III discusses the features needed of a BCI system in order to be useful for patients suffering from ALS. Section IV describes the characteristics of a humanoid robot that can be used for assistance in this context, analyzing a fundamental issue such as the energy efficiency, in comparison with human data. Finally, the expected results of the integration of these technologies and the future works are discussed.

II. PEOPLE SUFFERING FROM ALS: PROBLEMS, NEEDS AND STATE OF THE ART

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease involving the motor neurons. The disease has an incidence of ~ 1.9/100,000 and a prevalence of ~ 6/100,000.

Amyotrophic lateral sclerosis (ALS) is pathologically characterized by progressive loss of upper and lower motor neurons. The onset is usually from middle age to elderly, with a large variation in the presence of symptoms and evolution of the disease. The disease is often sporadic and its exact cause is not known, though an increasing number of genetic markers have been noticed recently. Even the

17th IEEE Mediterranean Electrotechnical Conference, Beirut, Lebanon, 13-16 April 2014.

387

disease duration is variable with a large spectrum of possibilities ranging from forms with mild conditions and a slower evolution to more severe forms with survival no more than 2.5 years and finally to very rapid forms with survival of less than 6 months.

Clinically the ALS is characterized by rapidly progressive weakness, muscle atrophy and fasciculations, muscle spasticity, difficulty of speaking (dysarthria), difficulty of swallowing (dysphagia), and difficulty of breathing (dyspnea). Sensory nerves and autonomic nervous system are generally unaffected, meaning the majority of people with ALS will maintain hearing, sight, touch, smell, and taste.

This progressive neurodegenerative disease, usually leads to complete immobilization of patients who need of a complete and continuous assistive support for living. In the majority of these patients, however, the cognitive functions are longer preserved making indispensable to find solutions that may help patients to contact caregivers and other people and to interact with the environment. The disease evolves toward a respiratory and cardiac failure, though death occurs after 3–5 years from the onset.

Since the time of Charcot’s first description of ALS between 1860-70, key discoveries in neurobiology have led to steady advances in our understanding of the pathogenesis of the disease and there has been modest progress in the management and reliable medical treatment of ALS. At the same time, the number of newly affected individuals has increased in the past few years as well as the identification of new familial forms. As no effective treatment is so far available for ALS, most medical interventions aim to improve the quality of life of patients.

Indeed, there are many opportunities for interventions that may improve quality of life for the patient and caregiver. Although the majority of the efforts are directed toward the individuation of tools that may maintain the capacity of communicating with family members and other people, other assistive aspect cannot be underestimated. In particular those aspects related to depression, hopelessness, anxiety, and other mental health issues that should be aggressively addressed and treated. Additionally many symptoms such as pseudobulbar palsy, scialorrhea, constipation, spasticity, and cramps can be treated effectively with medications. In contrast to the discouraging view that "there is nothing we can do," a broad approach to management, through collaboration with a multidisciplinary team, will permit the ALS physician to make a meaningful difference in the lives of individuals living with ALS. Table 1. Amyotrophic Lateral Sclerosis: Clinical Presentation and Prognostic Aspects

limb onset 75% limb weakness slower evolution

bulbar onset 25% Problem of speaking and swallowing

rapid progression

Respiratory onset Intercostals muscle weakness

rapid progression

fronto-temporal dementia onset

mood and cognitive changes

mild/rapid progression

In the context of the project BRAINHURO, Braincontrol transforms thoughts into action using the electrical activity generated by ensembles of cortical neurons. This has been made possible due to advances in methods of electroencephalography (EEG) analysis and in information technology, associated with a better understanding of the functional significance of certain EEG parameters. This system is particularly suitable for patients who have lost all motor function because is driven purely by thoughts and does not require sight, facial movements or any other stimuli.

III. BCI TECHNOLOGIES AND HUMAN-ROBOT INTERACTION

The main task of a BCI is the capability to distinguish different patterns of brain activity, each being associated with a specific intention or mental task. Most BCIs rely on non-invasive electroencephalogram (EEG) signals, where the electrical brain activity is recorded from electrodes placed on the scalp, that is a practical way to bring BCI technology to a large population.

Despite progress in Assistive Technologies (AT), there is still a large number of people with severe motor disabilities who cannot fully benefit from AT due to their limited access to current assistive products. For example, independent mobility is central to being able to perform activities of daily living by oneself. However, power wheelchairs are not an option for many people who, due to severe motor disabilities, are unable to use conventional controls. For some of these people, non-invasive brain-computer interfaces (BCIs) offer a promising solution to this interaction problem.

Over the past 20 years, research in the field of brain-computer interface has been increasing almost exponentially, driven by advances in understanding of brain function and the evolution of computers and sensors.

BCI is an effective way to augment human capabilities by providing a new interaction link with the outside world and is particularly relevant as an aid for disabled people.

In recent years, BCI research is exploring many applications in different fields related to AT, including communication [29], environmental control [30], and robotics and mobility [31-35].

However, no sufficiently usable and robust BCI solution addresses the needs of people with severe physical disabilities. BrainControl (see Figure 1) is the world’s first commercial available BCI that is usable by people who cannot move any muscles or communicate, but who are consciously aware, a state called “locked-in” or “apparent coma”. It is a BCI platform that allows people suffering from pathologies such as Amyotrophic Lateral Sclerosis (ALS), Multiple Sclerosis, tetraplegia and various kinds of muscular dystrophies, to overcome severe physical and communicative disabilities (3.7 million people worldwide). In particular, BrainControl can help patients suffering from diseases that paralyze the whole body or parts of the body, but who retain their intellectual abilities.

The first version of BrainControl, the “Basic Communicator”, fills a technological void for patients in

17th IEEE Mediterranean Electrotechnical Conference, Beirut, Lebanon, 13-16 April 2014.

388

“locked-in” state, and meets many of the unmet needs for patients in less advanced states who are currently using or cannot use eye-tracking systems.

Future versions of BrainControl, which are currently under development, will include advanced communication and entertainment (virtual keyboard, text-to-speech, web browsing, interaction via SMS, social networks, email, web-radio), home automation (lights, alarms, temperature, etc.) and the control of a wheelchair.

The aim of this work is to apply the BrainControl technology to the control of a humanoid robot for domestic assistance developed by the innovative Italian start-up Humanots.r.l. [36].

Figure 1. BrainControl.

IV. ENERGY EFFICIENCY: HUMAN-ROBOT COMPARISON A humanoid robot that interacts with the human

environment such as a home, a hospital, a public place, etc., should have enormous dexterity to avoid obstacles such as chairs, beds, tables, people, etc.. The dexterity is related to the degrees of freedom (DoFs) of the robot and to the joints motion controllers of the biped structure. Higher dexterity usually implies that many joints should be controlled and a big amount of energy is necessary. Optimising a robot structure, high dexterity with low energy consumption could be obtained. In the assistive application proposed with the BrainHuro project, the robot will be constantly at the ready, making energy efficiency a focal point of its implementation. In order to measure the power consumption and to compare data with the human gait and investigate the best suited regions of the body for energy recovering, a series of experiments were performed with the humanoid robot SABIAN (Sant’Anna BIped humANoid) [26], a biped humanoid robot developed by the Robot-An Laboratory, at Scuola Superiore Sant’Anna. It is a copy of WABIAN 2-R (Waseda BIped humANoid) [27], [28]. Compared to most bipedal humanoid robots, which walk with bent knees, WABIAN 2-R is able to perform a human-like walking, with stretched knees, and to get the pelvis motion, raising the hip. WABIAN 2-R is approximately the size of the average adult Japanese women. The robot has 7 DOFs in each leg, 2 DOFs in the waist, which help the robot perform stretched knee walking, 2 DOFs in the trunk. Every degree of freedom has a bioinspired range of motion, defined in reference to human motion measurements.

As it will be in the final version of the robot that will be used in this context, the iCub head was mounted on the

SABIAN body [26].

Two different gaits (knee-stretched and knee-bended) were performed on SABIAN with the aim of measuring the power consumption of every single joint in different configurations, also in comparison with data related to human movements, in order to give the bases for the realization of an energy recovering system.

From the analysis of the shown graphs (see Figures 2, 3, 4, 5 and 6), obtained by the driver interface of the robot and by neuroscience literature, some considerations can be derived:

1. The main percentage of the power is necessary to move the right and left knee pitches.

2. As well, the others legs pitches related to the ankle and the hip can play an important role in an energy recovering system.

3. Other joints are almost not used in this gait.

Considering the mechanical power in absolute value (see Figure 2), we can have an idea of the total amount of power necessary to perform a 4 steps knee-stretched gait [27].

It can be immediately noticed that the total amount of power during a simple walk is quite low, with peaks of 400 W (P_tot graph in the Figure 2). A big amount of power (30% - 50%) is thermally dissipated (electronic drivers, motor resistance); and it can hardly be recovered, although it can be reduced with measures in the design phase (P_th graph in the Figure 2). Some of the mechanical power can instead be recovered by using alternators instead of fixed bearings for belt tensioning (P_mecc graph in the Figure 2).

From an accurate analysis of the gait, we can mark different joint power regions (see Figure 3): stance flexion, stance extension, pressing, swing flexion, swing extension. During stance flexion, the motors that act to extend the knee are active, producing an extensor moment. The knee is flexing as the leg accepts the weight of the rest of the body, resulting in negative joint power. A negative joint power is also produced during the swing flexion phase, due to extensor knee moment.

Figure 2. Total amount of power dissipated during a knee-stretched

forward gait by SABIAN robot.

Regions of high joint power are best viewed as potential regions for energy recovering. As we can see, the swing

17th IEEE Mediterranean Electrotechnical Conference, Beirut, Lebanon, 13-16 April 2014.

389

extension phase is the region in which the largest amount of joint power is performed by the knee pitch motor, while also in stance flexion and in pre-swing phases a huge amount of power is required. For the other pitch joints, instead, stance flexion and stance extension are the movement phases in which we can focus in order to recover energy.

An interesting comparison can be done with the human walking, showing that the SABIAN platform [26-28] walks in a way that is very close to the human one. Figure 4 shows the behavior of the human knee hip joint during a streched-knee walk and the behavior of the same joint in the SABIAN robot: since the pattern show a very good similarity, strategies close to the one designed to implement energy recovering systems during a human walk can be here applied to the robotic platform.

Almost the same considerations made for the knee-stretched mode can be repeated of a knee-bended walking motion. The data are related to a gait of 4 steps with the knee bended with a degree of -50 degrees (where 0 degree is the stretched position).

Figure 3. Graphs related to the ankle (left), knee (center) and hip (right)

pitch joints of the right leg during a knee-stretched forward gait.

Figure 4. Comparison between knee data in humans and in SABIAN in a knee-stretched forward gait. Data related to the human gait are from

[37].

In figure 5, we can see high spikes of thermic power related to knee joint, due to a high amount of torque necessary for a very small period (this can be due to a not perfect behavior of the related harmonic drive). Again, this problem will be taken into account in the design phase. Considering the mechanical power in absolute value, we can have an idea of the total amount of power necessary to perform the 4 steps knee bended gait.

Finally, from the analysis of the gait of Figure 6, we can select here similar joint power regions in order to implement energy recovering systems.

Figure 5. Total amount of power dissipated during a knee-bended forward gait.

Figure 6. Energy consumption data during a knee-bended forward gait for the ankle (left), knee (center) and hip (right) pitch joints.

Anyway, as it can be easily seen, both in case of bent-knee or stretched-knee walking, the energy consumption is quite reduced, with pattern similar to the human one and some region of high joint power, on the pitch joints of both legs, that can be easily used for some strategy of energy recovering, increasing the autonomy of these robotic artifacts.

17th IEEE Mediterranean Electrotechnical Conference, Beirut, Lebanon, 13-16 April 2014.

390

CONCLUSION AND FUTURE WORKS This paper analyses the adoption of humanoid robotics

and BCI technologies to help people suffering from amyotrophic lateral sclerosis (ALS). The improvement of their quality of life is a fundamental aspect and it was shown that the possibility to interact with robotic artifacts could be very useful at this aim. This possibility is currently being investigated in the context of the project BRAINHURO, in which the techniques of BCI systems are integrated with a humanoid robot in order to assist ALS patients. First data related to the energy consumption of the humanoid platform that will be used in this context are shown in the paper. Next work will be oriented to show the results of the integration of BCI technologies here addressed, both considering engineering and a psychological point of view.

ACKNOWLEDGMENTS This joint research was partly supported by the

BrainHuRo Italian Project (cofounded by the POR-CREO Fesr 2007-2013 – Le ali alle tue idee) and partly supported by University of Siena, Liquidweb s.r.l. and Humanot s.r.l. as active partners in the BrainHuRo project.

REFERENCES [1] Stoianovici, D. "Robotic surgery". World journal of urology.18.4

(2000): 289-295. [2] Agarwal, R., Levinson, A. W., Allaf, M., Makarov, D. V., Nason, A.,

& Su, L.-M. (2007), “The RoboConsultant:Telementoring and Remote Presence in the Operating Room During Minimally Invasive Urologic Surgeries Using a Novel Mobile Robotic Interface”, Urology, 70:5, pp. 970-974.

[3] Harris, S. J., Arambula-Cosio, F., Mei, Q., Hibberd, R. D., Davies, B. L., Wickham, J. E. A., Nathan, M. S., Kundu, B. (1997). “The Probot – an active robot for prostate resection”. Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine, 211:4, pp. 317-325.

[4] F. E. Cepolina, M. Zoppi, “Design of multi-degrees-of-freedom dexterous modular arm instruments for minimally invasive surgery”, Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in MedicineNovember 2012 vol. 226 no. 11827-837.

[5] Sung, Gyung Tak, and Inderbir S. Gill. "Robotic laparoscopic surgery: a comparison of the da Vinci and Zeus systems". Urology, 58.6 (2001): 893-898

[6] Adler Jr., J. R., Chang, S. D., Murphy, M. J., Doty, J., Geis, P., & Hancock, S. L. (1997). “The Cyberknife: a frameless robotic system for radiosurgery”. Stereotact Funct Neurosurg, 69: 1-4, Pt. 2, pp.124-8.

[7] Drexler, K. E. (2004). “Nanotechnology: from Feynman to Funding”. Bulletin of Science, Technology and Society, 24, pp. 21-27.

[8] Courreges, F., Vieyres, P., Poisson, G. (2008). “Robotized tele-echography”. In S. Kumar, & E.A. Krupinski(eds.), Teleradiology (pp. 139-154), Berlin Heidelberg: Springer.

[9] Susilo, Ekawahyu, Pietro Valdastri, Arianna Menciassi, and Paolo Dario. "A miniaturized wireless control platform for robotic capsular endoscopy using advanced pseudokernel approach". Sensors and Actuators A: Physical156, no. 1 (2009): 49-58.

[10] Connolly, Christine. "Prosthetic hands from touch bionics", Industrial Robot: An International Journal35.4 (2008): 290-293.

[11] Cipriani, Christian, Marco Controzzi, and M. Chiara Carrozza. "Progress towards the development of the SmartHand transradial prosthesis". IEEE International Conference onRehabilitation Robotics, 2009. ICORR 2009.. IEEE, 2009.

[12] Veneman, Jan F., et al. "Design and evaluation of the LOPES exoskeleton robot for interactive gait rehabilitation". IEEE Transactions onNeural Systems and Rehabilitation Engineering, 15.3 (2007): 379-386.

[13] Hogan, N., et al. "MIT-MANUS: a workstation for manual therapy and training".Proceedings., IEEE International Workshop on Robot and Human Communication, 1992.. IEEE, 1992.

[14] Kidd, Cory D., Will Taggart, and Sherry Turkle. "A sociable robot to encourage social interaction among the elderly".IEEE International Conference onRobotics and Automation, 2006. ICRA 2006. Proceedings 2006, pp. 3972-3976. IEEE, 2006.

[15] Pioggia, G., Ahluwalia, A., Carpi, F., Marchetti, A., Ferro, M., Rocchia, W., & De Rossi, D. (2004). “FACE: Facial Automaton for Conveying Emotions”. Applied Bionics and Biomechanics, 1:2, pp. 91-100.

[16] G. G. Muscolo, C. T. Recchiuto, G. Campatelli, R. Molfino, “A robotic social reciprocity in children with Autism Spectrum Disorder”, G. Herrmann et al. (Eds.): ICSR 2013, LNAI 8239, pp. 574--575. Springer International Publishing Switzerland (2013).

[17] Powell, S. "The use of computers in teaching people with autism". In Autism on the agenda: papers from a National Autistic Society Conference. London. 1996.

[18] Dautenhahn, Kerstin. "Robots as social actors: Aurora and the case of autism". In Proc. CT99, The Third International Cognitive Technology Conference, August, San Francisco, vol. 359, p. 374. 1999.

[19] Shamsuddin, Syamimi, Hanafiah Yussof, Luthffi Ismail, Fazah Akhtar Hanapiah, Salina Mohamed, Hanizah Ali Piah, and Nur Ismarrubie Zahari. "Initial response of autistic children in human-robot interaction therapy with humanoid robot NAO". In Signal Processing and its Applications (CSPA), 2012 IEEE 8th International Colloquium on, pp. 188-193. IEEE, 2012.

[20] Kurt, Anja, Femke Nijboer, Tamara Matuz, and Andrea Kübler. "Depression and anxiety in individuals with amyotrophic lateral sclerosis".CNS drugs21, no. 4 (2007): 279-291.

[21] Lebedev, M. A., & Nicolelis, M. A. L. (2006). “Brain-machine interfaces: past, present and future”. Trends in Neurosciences 29:9, pp. 536-546

[22] Hochberg, L. R., Serruya, M. D., Friehs, G. M. et al. (2006). “Neuronal ensemble control of prosthetic devices by a human with tetraplegia”. Nature 442, pp. 164-171.

[23] Galán, Ferran, Marnix Nuttin, Eileen Lew, Pierre W. Ferrez, Gerolf Vanacker, Johan Philips, and J. del R. Millán. "A brain-actuated wheelchair: asynchronous and non-invasive brain–computer interfaces for continuous control of robots".Clinical Neurophysiology119, no. 9 (2008): 2159-2169.

[24] Asfour, T., Regenstein, K., Azad, P., Schröder, J., Vahrenkamp, N. & Dillmann, R., "ARMAR-III: An Integrated Humanoid Platform for Sensory-Motor Control", In IEEE/RAS International Conference on Humanoid Robots (Humanoids). Genova, Italy. December 2006., pp. 169-175.

[25] E. Falotico, D. Zambrano, G. G. Muscolo, L. Marazzato, C. Laschi, P. Dario, “Implementation of a bio-inspired visual tracking model on the iCub robot”, Proceedings of the 19th IEEE International Symposium on Robot and Human Interactive Communication – 12-15 September 2010, Viareggio, Italy.

[26] G. G. Muscolo, C. T. Recchiuto, K. Hashimoto, P. Dario, A. Takanishi, “Towards an Improvement of the SABIAN Humanoid Robot: from Design to Optimisation”, Journal of Mechanical Engineering and Automation, Scientific & Academic Publishing, Vol.2, No.4, 80-84, July 2012.

[27] Muscolo, G.G.; Recchiuto, C.T.; Laschi, C.; Dario, P.; Hashimoto, K.; Takanishi, A.; "A method for the calculation of the effective Center of Mass of humanoid robots", 11th IEEE-RAS International Conference onHumanoid Robots (Humanoids 2011), 2011, pp.371-376, 26-28 Oct. 2011, Bled - Slovenia

[28] G. G. Muscolo, K. Hashimoto, A. Takanishi, P. Dario, “A Comparison between two force-position controllers simulated on a humanoid arms”, Journal of Robotics, Hindawi Publishing Corporation, 14 pages, Volume 2013.

[29] B. Obermaier, G. R. Müller, ““Virtual Keyboard” Controlled bySpontaneous EEG Activity”, IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING , VOL. 11, NO. 4, DECEMBER 2003.

[30] Elizabeth A. Bayliss, John F. Steiner, Douglas H. Fernald, Lori A. Crane, Deborah S. Main, “Descriptions of Barriers toSelf-Care by Persons withComorbid Chronic Diseases”, ANNALS OF FAMILY

17th IEEE Mediterranean Electrotechnical Conference, Beirut, Lebanon, 13-16 April 2014.

391

MEDICINE,WWW.ANNFAMMED.ORG, VOL. 1, NO. 1, MAY/JUNE 2003.

[31] José del R. Millán and Josep Mouriño, “Asynchronous BCI and Local Neural Classifiers:An Overview of the Adaptive Brain Interface Project”, IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, VOL. 11, NO. 2, JUNE 2003.

[32] Miguel A. L. Nicolelis and John K. Chapin, “Controlling Robots with the Mind”, Scientific American:, September 16, 2002. 1996-2005 Scientific American, Inc..

[33] Carmena JM, Lebedev MA, Crist RE, O'Doherty JE, Santucci DM, “Learning to Control a Brain–Machine Interface for Reaching and Grasping by Primates”. PLoS Biol 1(2): e42.doi:10.1371/journal.pbio.0000042.

[34] José del R. Millán, Frédéric Renkens, Josep Mouriño, “Noninvasive Brain-Actuated Control of a MobileRobot by Human EEG”, IEEE

TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 51, NO. 6, JUNE 2004.

[35] MICHAEL S. LEW, NICU SEBE, CHABANE DJERABA, RAMESH JAIN, “Content-Based Multimedia Information Retrieval:State of the Art and Challenges”, ACM Transactions on Multimedia Computing, Communications and Applications, Vol. 2, No. 1, February 2006, Pages 1–19.

[36] http://www.humanot.it [37] Li, Qingguo, Veronica Naing, J. Andy Hoffer, Doug J. Weber, Arthur

D. Kuo, and J. Max Donelan. "Biomechanical energy harvesting: Apparatus and method." In Robotics and Automation, 2008. ICRA 2008. IEEE International Conference on, pp. 3672-3677. IEEE, 2008.

17th IEEE Mediterranean Electrotechnical Conference, Beirut, Lebanon, 13-16 April 2014.

392