Smart Kitchen (Ambient Assisted Living) Ameya Daphalapurkar 17 January 2014, www.mdpi.com/journal/sensors ISSN 1424-8220 Article A Smart Kitchen for Ambient Assisted Living Rubn Blasco 1,*, lvaro Marco 1, Roberto Casas 1, Diego Cirujano 1 and Richard Picking 2 Overview
Introduction Related Work System Description
Software Architecture System Evaluation Conclusions and Comments AAL Ambient Assisted Living (AAL) are concepts, products and services which combine new technologies and the social environment in order to improve quality of life in all periods of life.
Problem - Ageing of the population Ratio of people aged 65 or more will increase to 30.0% in 2060 in Europe
20.2% in 2050 in USA 39.6% in 2050 in Japan Issues Physical and/or Cognitive Impairments as age increases Reduced speed and increased time to make precise movements Affects sensing and information processing capability
Difficulties in multi tasking Loss in capabilities to autonomously perform activities Need for Safety Old people most vulnerable to domestic accidents Most domestic injuries are related to working in kitchen Harm from kitchen tools, cutlery and household appliances
Consequence : may decide moving to nursing home About the paper Easy Line+
Increasing elderly and disabled peoples autonomy The kitchen is the focus Many activities that are key for autonomy are performed Ambient Intelligence Ambient Intelligence (AmI) can be defined as a sensitive and adaptive electronic environment that responds to the actions of the persons and object and cater for their needs
This approach includes the entire environment, taking into account each individual object, associating its interaction with humans AAL uses the AmI as the essential tool to provide integral solutions for supporting the person in his/her independent living in different contexts: dwellings, transport, workplaces, etc. The three macros for AAL AAL4persons ageing well for the person
AAL in the community social inclusion [email protected] elderly and people with disabilities at work Related Work MONAmI project selects suites of technological services to support people at risk of exclusion and loss of autonomy Necessity system proposed by Muoz et al. which offers a system to represent and validate alerts in a domestic environment
Lei show a system based only in a RGB-D camera (modern depth cameras that provide synchronized color and depth information at high frame rates) which identifies activity and tools used from a set of objects and actions Related Work Suryadevara and Mukhopadhyay developed and tested an intelligent home monitoring system based on a wireless sensors network (no camera
or vision sensors) to monitor and evaluate the well-being of the elderly Ficocelli and Goldie present an assistive kitchen with speech communication and an automated cabinet system to ease storing and retrieving items and to obtain recipes for meal preparation. Schwartze present their work in graphical interfaces for Smart
Environments with the 4-star Cooking Assistant application which proves the capability of their system to dynamically adapt a graphical user interface to the current context of use System Description Four main functionalities within the kitchen scenario: 1. Facilitates the use of household appliances 2. It provides useful information and warnings 3. It detects emergency situations and takes corrective
actions 4. It analyzes all the data gathered to extract relevant information Principles Two main principles guided the system design: 1. Resistance to obsolescence 2. Ability to interoperate with existing systems in the field
(such as white goods, sensors or RFID from different manufacturers) Requirements Need to reduce unitary price and complicate installations A central intelligence entity It is conceived as a set of interchangeable blocks with defined communication systems
interfaces to grant interoperability
among existing Any electrical appliance with communication capability can be integrated. As a result, the development and stability of the appliances eases, they dont change their current way of functioning, but they just need to add communication to inform about their status and execute actions
Integrations for the system Power Line Communication (PLC) for the white goods
RFID for item identification ZigBee as wireless sensor network Infrared for the remote control Bluetooth for audio streaming Ethernet (WiFi and cable) for cloud and user interaction Fig: Smart Kitchen Context Interaction
RFID with ZigBee communication A stand-alone RFID reader in worktop to gather any information Patching label including RFID chip and metalized thread technology Not so standard sensors with ZigBee: Magnetic, Light, Presence
Human-Machine Interface (HMI) devices Mobile, Tactile, Embedded devices Fig: Communication diagram for context interaction in the Smart Kitchen E-Servant System intelligence is provided by the e-Servant
Learning system, which detects and compensates the behavior, habit changes and loss of abilities of the user. Checks continuously the state of the kitchen appliances, providing warnings through its user interfaces if there is any problem or event to be notified Detects emergency situations and takes corrective actions Also manages records with the relevant events that have occurred in
the kitchen gathered from the context and user interaction E-Servant Data is processed and analyzed in order to extract findings about the cognitive level of the person that could be useful to the guardian and/or relatives. Information is used to create Quality of Life Evaluation (QoLE):
1. A detailed report food management, cooking, doing laundry 2. Suggestion on the support level of the system Software Architecture Designs based in Service Oriented Architectures (SOA) SOA technology - Open Services Gateway initiative (OSGi)
Pieces of code are organized into bundles that can be managed separately. Handling a serial port, providing a command line interface, collecting, aggregating and analyzing data, etc. OSGi Manages these bundles dynamically Providing new features and capabilities by adding new
services Backbone of the e-Servant in order to enhance its capabilities, and decreasing the cost of maintenance in a future Fig: Software architecture of the e-Servant Context Manager
Information about the status of the appliances, product inventory, user actions or any other event is gathered by the CM and sent to the Logic Unit (LU) which will decide whichever operation must be performed The CM is the agent responsible for retrieving that information, processing and presenting it in a structured way, and it is organized in three levels: - Drivers
- Devices - Devices management Driver Lowest layer of CM which communicates with physical devices Three important tasks: 1. Physical channel establishment
2. Device enumeration and network support 3. Device installation and messaging service PLC driver, ZigBee driver Devices Devices maintain a link with the driver which has instantiated them, and OSGi provides the mechanism to dynamically modify this link if the base driver disappears (for example, if a network gateway becomes unavailable)
Device Manager The Device Manager is responsible for manipulating and aggregating information from the devices and effectively offering the context awareness to the upper layers 1. Database logging 2. Action driving 3. Event triggering
Fig: Architecture of Context Manager Logic Unit Brain of the e-Servant Important tasks: 1. Process all the information provided by the context manager 2. Reason through that information and deciding actions in order to support the user
3. Cooperate with the User Controller Interface (UIC) to manage the interaction with the user from a logical perspective Fig: Architecture detail of the Logic Unit Fig: User interface screen showing information about the washing machine status
Quality of Life Evaluation System The Quality of Life Evaluation System is a service that periodically (a period configurable between 1 and 3 months) analyses the context database looking for changes in the user washing, shopping and cooking habits which could be relevant in order to detect a loss of physical, cognitive or sensorial capabilities Example, if the user starts going to the fridge at night (might indicate
insomnia) or if s/he is doing the laundry less and less often (might indicate that he/she is wearing dirty clothes). This allows performing an indirect evaluation of the quality of life of the user Designed for the use of non-technical people Use Case Smoke sensors notify the system that there is smoke in the kitchen, oven and hob are on but nobody is in the kitchen.
Fig: Use Case Scenario 1 to 5 Fig: Use Case Scenario 6 to 10 System Evaluation The system has been evaluated by 63 end users and 31 formal and informal carers in two living labs placed in Spain and UK
Each user evaluates the system through four specific situations There are three people participating in the assessment whose roles are: - The user is the person who will evaluate the technology - The test moderator who leads the sessions - The test observer who is watching the different situations evaluated without contact with the user Fig: Evaluation Process
Evaluation Results In summary, we can say that: The system has good usability and physical, sensory and cognitive accessibility 90% of the users that evaluated the system found it accessible
Usability has been evaluated with a score of 3.85 out of 5 overall, on a rating scale of 1 (poor) to 5 (excellent). Fig: Functionalities of the e-Servant evaluated by the caregivers and users Conclusions The system concept and its implementation are innovative
The backbone of the system is its modular architecture based on an OSGi framework Functionalities of the system can be easily expanded by adding rules and user-scenarios Quality of Life Evaluation Service allows progressive personalization of the system
Comments on the paper PROS: - A great innovative concept with never ending possibilities for improvement and advancements. - Very well explained architecture and software organization description - Modularity in bundles clearly paves the way for ease in updates for the system without causing any huge change in the functioning of the system.
CONS: - I think the usability part could have been better explained by the authors especially as the target demographic is the elderly people and we should understand their grasp towards technology and user interfaces THANK YOU
Birth gametes fertilization zygote morula blastocyst implantation gastrula neurulation Heart beat EEG activity foetus > 24 weeks birth / neonate If there is a critical tipping point in foetal development, it is during the process of delivery of the neonate....
What is Associative Learning (AL) and why do we need it? History of AL implementation in ACT-R. Bayesian log-likelihood transformations. From a Bayesian to Hebbian Implementation. Recent Neural Evidence: Spike Timing Dependent Plasticity. A balanced associative learning mechanism. Hebbian and...
The persona describes red as a sunset which is bright and blazing because of its colour. Red gives a feeling of bravery that comes within you. Red is described as a sunburn spot which appears in red; caused through the...