Towards an Intelligent Service to Elders Mobility Using the i-Walker

The SHARE-it architecture is designed to inform and assist the user and his/her caregivers through monitoring and mobility help. Thus, we plan to contribute to the development of the next generation of assistive devices for older persons or people with disabilities so that they can be self-dependent as long as possible. In this paper we focus on the development of an intelligent pedestrian mobility aid that we call ı-Walker. ı-Walker is a robotically augmented rollator to reduce fall risk and confusion, and to increase rollator convenience and enjoyment. SHARE-it provides an Agent-based Intelligent Decision Support System to support elders with cognitive and/or motorial problems. The main goal of the SHARE-it (SHAREit 2007), an EU FP6 funded project, is to contribute to the development of the next generation of intelligent and semi-autonomous assistive devices for older persons and people with disabilities (both cognitive and/or motor) so that they can be selfdependent enough to autonomously live in the community, staying at home as long as possible with a maximum safety and comfort; this possibility would increase their quality of life, and, at the same time, delay their institutionalization. At least in part, how well these elders live and their abilities to maintain independent life styles will depend on their health and the degree to which they have remained totally able or frail or disabled. When we talk on frailty and disability we refer to the most recent definition, in particular frailty is described as: A state of increased vulnerability to stressors that results from decreased physiological reserves and multi-system dysregulation, limited capacity to maintain homeostasis and to respond to internal and external stresses. Frailty is an aggregate expression of risk resulting from ageor disease-associated physiologic accumulation of subthreshold decrements affecting multiple physiologic systems resulting in adverse health outcomes (Fried et al. 2004). Disability is defined as the difficulty or dependency in carrying out activities necessary for independent living, including roles, tasks needed for self-care and household chores, and other activities important for a persons quality of life (Fried et al. 2004). Copyright c © 2008, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. This in turn, will depend to at least some extent on how well the artificial and built environments in which they live conform to their needs and their age-related losses in abilities and somatic integrity. There are consequently two important targets in future research focused on ageing population: • to develop consistent and valid methods for assessing frailty and disability in order to individuate the main areas in which this target population needs help; • to design environments appropriate for the proportion of elders who are increasingly frail and disabled but whose Life Expectancy expectancy continues to grow Different groups of elders are characterized by large variations in health, well-being, disability, and health care needs. Since Assistive Technologies (AT) major purpose is represented by the possibility of assisting people elderly and/or disabled to stay or to get back to their home, a target population in terms of functional, social, and clinical features has to be individuated so that a number of individuals as large as possible can achieve benefits as effectively as possible (Mittal et al. 1998). According to these premises, SHARE-it Target Population will operatively be individuated through the assessment of the presence of: • Mild functional impairment: older and/or disabled populations are made up by individuals who present widely different and heterogeneous functional profiles. Considering themore prevalent diagnosis in this group of persons (stroke, Parkinsons disease, Alzheimers disease), impairments range from extremely mild (people able to walk with a cane or affected by such a mild memory loss that allows them to live on their own) to extremely severe (persons bedridden or completely unable to understand a simple order). In the first case we are in the realm of prevention, in the second in that of palliative care or institutionalization. People with mild functional impairment those in the middle area of impairment are expected to have the best results through the use of proper assistive technology and are the target of the assistive intervention. • Possibility of changing functional profile: mild disability is characterized by the possibility of increasing the functional level according to proper assistive or rehabilitative interventions. Users of a flexible assistive device can benefit from its adaptation so that it could help to improve their condition as a result of rehabilitation treatment, or can prevent the decrease of their autonomy due to intervening conditions or diseases. This paper considers the definition of SHARE-it users and those Activities of Daily Living (ADL) where SHARE-it hardware devices can offer some kind of personalized support to them. The considered elderly population is assessed through the multidimensional geriatric approach. Users are classified on the basis of their functional impairment as suggested by World Health Organization (WHO). Furthermore, since one of the parameters that affect the user’s performance in a specific task is the emotional reaction, we are considering the alarm reaction as an aspect to be detected and to be used as the basis on the elaboration of the decision to give or not support to the user. Also, it will be used to decide which kind of support to give to each individual. Many older adults use walkers to improve their stability and safety while walking. A walker, may support up to 50% of the users body weight, are ideal for weak knees or ankles or severe balance problems. Also, if some individuals, in this population, do not need to lean on their walker for balance, they might be able to walk faster with the aid of a walker and then they might improve their autonomy. However, there are many subjects that can not use a walker for risks related to a particular mobility impairments or at the presence of a cognitive disorders. The traditional walker is prevented for people who suffered from lack of strength in their arms or legs (tipically subjects affected by stroke, a cardiovascular impairment that, as mobility damage, can affect half part of body). The correct instruction to use a walker is to move pushing the device but this situation is not safe in this population because the different degree of strength in two sides of walker create an uncorrected trajectory, a balance problem and an high risk of fall. About cognitive problems difficulties in orientation is one of the most frequent reason to prevent the use of walker. In order to help this population the same target population of SHARE-it we have developed a robotically augmented rollator. This device is aimed to reduce fall risk and confusion, and to increase users convenience and enjoyment and finally and most important aspect to increase autonomy in persons prevented to use a traditional walker. ı-Walker is designed to be situated in the user’s preferred environment (?). This strong constrain implies both that the user knows this space – s/he lives there – and, also, that the SHARE-it agent-based platform is deployed in this environment (Barrué et al. 2005). Thus the ı-Walker is situated and knows this space. An important lesson learned while developing ı-Walker is that the elderly population requires techniques that can cope with individual differences. We choose intelligent agents as the adaptation element. Among of the SHARE-it objectives is to build different ıWalker workbench platforms, oriented to demonstrate their feasibility, and gain the confidence to support the specific disabilities (Cortés et al. 2003). Two inspiring works in this line for intelligent pedestrian aids are (Glover et al. 2003) and (Wasson et al. 2001; 2003). Those propose intelligent robotics agents to support elders mobility but the main difference in our approach is the explicit use of intelligent software agents to support decision-making and to help in the interfacing with the user. More modern lines of research in the field of pedestrian mobility are represented by (Cheng, Bateni, and Maki 2008) and (Kulyukin et al. 2008). The first is more oriented to avoid the fall risk and the second culminates a line of walkers developed at CMU. The Kulyukin et al’s iWalker is intended to guide elders in an Small World (SW) and it is not prepared to avoid unexpected obstacles in that environment. Other walkers incorporate robotic aids that are prepared to help users to change their position as from sit-to-stand (?). Standing up motion is one of the most serious and important operations in daily life for elderly person who does not have enough physical strength. An important issue to be considered is that before starting experimentswith elders the whole system has to be approved by a Ethical Committee. We had use the original agentbased control elements in an experiment with volunteer inpatients in Fondazione Santa Lucia, Rome, using Spherik an intelligent wheelchair (Annicchiarico et al. 2007). In this paper, we generalize the agent-based system to be used in the ı-Walker. Although, the experimentation with elders has to start the whole system is already in place in Casa Agevole (Vescovo 2005). Autonomy and Disability According to prevailing models of the disablement process, disability results when these diseases and conditions, via specific impairments and functional limitations, lead to limitations in the ability to perform basic social roles. Hence, disability is usually defined as the degree of difficulty or inability to independently perform basic activities of daily living (ADL) or other tasks essential for independent living, without assistance. In order to quantify residu