A cerebral vascular accident, known as common language stroke, is one of the main causes of mortality and remains the primary cause of acquired disabilities in adults. Those disabled people spend most of their time at home in their living rooms. In most cases, appliances of a living room (TV, light, cooler/heater, window blinds, etc.) are generally controlled by direct manipulation of a set of remote controls. Handling many remote controls can be disturbing and inappropriate for these people. In addition, in many cases these people could be alone at home and must open the door for visitors after their identification by either moving to the door or using an intercom system which requires in both cases a physical activity. Furthermore, these people need a continuous health monitoring especially blood pressure to avoid a recurrent stroke. Smart spaces and assisted technologies would be beneficial to assist person with disabilities to live independently, enhance their quality of life and empower their autonomy. A complete system which improves and facilitates the daily life and covers all aspects such as appliances automation, appropriate interaction mode and health monitoring of these people is still lacking. The aim of this work is to create a safe and high-quality living environment for persons with disabilities to enable them to live more independently by automating the operation of a living room appliances according to the current context without the need to use remote control devices, the use of a suitable interaction modality with appliances that require direct interaction and a remote health monitoring system which can alert relatives and caregivers in case of an emergency.
According to the World Stroke Organization (WSO), more than 13 million new stroke patients occur each year, among them 5 million die and 5 million are left permanently disabled [
Smart home or Ambient assisted living (AAL) technology is considered as the major application domain of IoT technology and could be an efficient tool to facilitate the improvement of autonomy, safety, and social participation of people with disabilities, particularly those living alone. AAL is a term that indicates the use of information and communication technologies to make the living environment adaptable for elderly or people with disability [
Very little work has been done in the literature for assisting people with disability caused by a stroke, except for a few attempts to promote independent living for a person with disabilities in general. Current research on disabled person assistance and AAL has mainly focused on two aspects: home automation and remote health monitoring. Home automation was achieved by learning the user’s daily routines to both detect the conformity of his behavior and foresee his next possible actions. By knowing the routine of the inhabitant, smart home can foresee what activities he/she is likely to do next, and when to adapt appliances accordingly or automate some daily routines. Such approach requires large data sets for training models. It is also difficult to apply modeling and learning results from one person to another. In addition, disabled person may resist having their activities monitored in any way. In addition, there may be a risk that video sequence used for daily routines learning is considered to be strongly privacy violating. Remote health monitoring was achieved either by identifying if the person needs a form of assistance since an unusual activity has been recognized or using wearable sensors to gather physiological and motion information that enables patient’s status monitoring [
In spite of the significant work proposed for both appliance automation and health monitoring there still a missing complete solution for disabled persons from stroke which embed the three major aspects of life assistance for these people namely context-aware appliances adaptation, appropriate modality interaction and health monitoring. The aim of this research paper is to propose a complete system that is a smart room to assist, and take care of, its inhabitants by proactively support people with disabilities caused by stroke in their daily activities and monitor their health to avoid a recurrent stroke. The proposed system consists of the following components: 1) a context-aware adaptation system able to adapt home appliances and devices according to the current context without explicit intervention from the inhabitant or use of remote-control devices; 2) a voice modality system to control devices that need explicit interaction from the inhabitant namely TV and intercom; and 3) a remote monitoring system for health control based on wearable sensors.
The remainder of this paper is organized as follows: Section 2 discusses related work. Section 3 specifies and analyzes the main components of the proposed smart room and outlines its architecture. Section 4 discusses implementation details of the smart room; Section 5 presents a short discussion of the proposed system and Section 6 concludes the paper and presents future work.
In the literature, various studies and systems aimed at monitoring and supporting people that need assistance have been proposed. Reference [
As stated before, most proposed solutions were not directly oriented to disabled persons due to stroke who needs home automation system, convenient way for interaction and a health monitoring system. Most of previous work proposed a home automation system either for normal person or disabled (resp. elderly) person without a health monitoring system or an appropriate interaction modality or just a health monitoring system. To the best to our knowledge a complete solution that encompass the aforementioned subsystems required for disabled person from stroke is still inexistant.
The smart room for a disabled person from a stroke is typically like a living room composed of usual appliances and devices. It is made up of three main subsystems namely (1) the interactive subsystem which operation depends on the actions of the person such as TV and intercom system; (2) the reactive subsystem which operates by reacting to stimuli from its environment, it is in continuous interaction with its environment such as the light system (set of light bulbs), heater and cooler system, air conditioner (resp. air purifier) and the window blinds; and (3) the health monitoring subsystem which monitor the blood pressure and heart beat rate of the patient, in addition to a mobile phone which ensure the communication of the measured bio-signals (
Some appliances of the smart room are reactive which means they are in continuous interaction with the surrounding contextual environment of the smart room and react accordingly. The reactive subsystem may be divided in two main components: the light system and the climate system. This subsystem has a layered architecture as indicated in Sensor layer: contains the basic sensors for raw context information capture especially indoor temperature and outdoor light brightness. This layer sends raw context to the context representation and interpretation layer. Context representation and interpretation layer: responsible of putting received raw context from the lower layer on a common format which can be understood by the upper layer. Generally, the format is a vector of values. In addition, the interpretation part of this layer performs interpretation of raw context to useful information such as converting the range of sensed temperature between −50 C and 50 C as a very could temperature (resp. freezing temperature). This layer sends the interpreted context in a suitable format to the adaptation layer. Adaptation layer: composed of two subsystems: 1) the light adaptation which perform adaptation of light inside the room according to the current context by controlling the light intensity of the set of light bulbs and the level of window blinds. 2) The climate subsystem which control the overall climate conditions in the room by controlling the operation of the cooler, heater, air conditioner and air purifier based on the current room temperature and degree of air purity. This layer takes as input the meaningful context vector from the context representation and interpretation layer and a set of command signals as output to devices and appliances layer. Devices and appliances layer contains a set of appliances and devices namely light bulbs, window blinds, cooler, heater, air conditioner and air purifier. All these devices, and appliances operates according to command signals received from the adaptation layer.
The light system is responsible of the luminosity adjustment inside the smart room using a set of light bulbs and the window blinds of the room. The lighting inside the smart room is adjusted by using the largest possible amount of external lighting from daylight to save energy consumption and turn on the lights in the absence of insufficient external lighting. This requires the window blinds to be opened at different levels according to the degree of external lighting, as well as the operation of lights with different lighting levels. This system will use a light sensor to the
Daylight degree | Window blinds | Light bulb luminosity |
---|---|---|
Dark | Closed | High |
Low | Totally opened | Average |
Average | Mostly opened | Low |
High | Mostly closed | Off |
In the same way the climate system which is composed of the heater/cooler and the air conditioner (resp. air purifier) will adjust the temperature inside the smart room as well as the air quality which could affect the health of stroke patients. This system will use both a temperature sensor and air quality sensor to adapt to the desired temperature and air quality level. The temperature automatically adjustment to an average temperature which fixed previously, and the air quality level is adjusted according to CO2 and smoke threshold.
Indoor temperature | Cooler | Heater |
---|---|---|
Very low | Off | High |
Low | Off | Average |
Almost low | Off | Low |
Average | Off | Off |
Almost high | Low | Off |
High | Average | Off |
Very high | High | Off |
Air quality | Air purifier |
---|---|
Clean | Off |
Very good | Very low |
Good | Low |
Average | Medium |
Bad | High |
Very bad | Very high |
There are two main devices in the smart room that could be considered as interactive systems which respond according to the patient actions: TV and intercom system. The patient needs to execute three basic operations on TV namely put TV ON/OFF, up and down the volume and channels. For this and to avoid fatigue of the patient by using manual manipulation of the TV remote control, an appropriate method of interaction based on voice will be used. It consists of associating the remote-control commands to voice words as shown in
Commands | Voice words |
---|---|
TV on | T1 |
TV off | T0 |
Volume+ | V plus |
Volume− | V minus |
Channel+ | C plus |
Channel− | C minus |
For the intercom system, the patient needs just to command the door opening using the voice modality to avoid moving to the door and open it manually. The door closure is done automatically without the need to command it. The patient communicates with visitors using the intercom monitor which should be place close to him. The voice word used to open the door is D1.
The general architecture of the interactive subsystem is given by
High blood pressure is the biggest cause of strokes especially when it is 140/90 or higher. Control of high blood pressure (hypertension) is one of the most important things we can do to reduce the risk of stroke. If a person had a previous stroke, it is possible to halve the risk of stroke recurrence, if blood pressure is constantly monitored. In our case the patient blood pressure is monitored using a wearable device (sensor) to continuously measure the blood pressure and communicate it to a mobile phone which send alerts to relatives or caregivers in case of emergency. In addition, the patient mobile phone will save the blood pressure measures for a certain period for further check and analysis (
The smart room is implemented using Arduino as a central hub. Arduino is open-source electronics prototyping platform based on flexible, easy-to-use hardware and software [
All the appliances (resp. devices) of the light subsystem are commanded using remote controls. The light bulb has two commands: ON and OFF, in addition to three level of luminance namely low (default on ON command), average and high. The window blinds have two commands: up and down. Whenever the user keeps pressing one of these buttons the store will continue the action either up or down depending on the pressed button. The subsystem will sense the outdoor light (daylight degree) using a light sensor and reacts accordingly to adjust the indoor light. This task will be done automatically without explicit intervention from the patient. To accomplish this task, a simple Infra-Red receiver is used to learn the different codes of the remote control for each command. This is done for both window blinds remote control and light bulb remote control. To learn the codes commands of the remote control we should place the remote control in front of the IR receiver and press the desired button; then using the appropriate Arduino IDE for IR remote controls we will get the hexadecimal code of the button.
As an application, the system has learnt the following IR commands of a TV remote control: FD00FF >>>> ON/OFF FD807F >>>> Vol+ FD906F >>>> Vol− FD50AF >>>> Ch+ FD10EF >>>> Ch−
The IR codes of basic commands for each remote control of the following devices had been learnt TV, air purifier, cooler/heater, window blinds, light bulbs, and door lock opening (via intercom system). All the learnt codes were saved in a table with the structure shown in
Device | Command | IR code |
---|---|---|
TV | On/off | FD00FF |
TV | Ch+ | FD50AF |
Light bulb | High | … |
Window blinds | Close (down) | … |
… | … | … |
The adaptation of the light system is based on the external lighting. For that reason, a light sensor (LDR: Light Dependent Resistor or simply photoresistor) is required to sense it. As mentioned before, the light system is composed of light bulb and window blind with the following states: Light bulb: Off On low On Average On High Window blind: Closed Totally opened. Mostly opened. Mostly closed.
The IR signals of the light bulb remote control are already learnt in the previous step. The window blind remote control has only two commands: down and up. To totally open (resp. close) the window blind, the down (resp. up) button should be continuously pressed during T seconds. For the state mostly opened (resp. mostly closed), the up (resp down) button should be continuously pressed during (T*3/4) (resp. T/4) seconds. An IR receiver was used for simulation purpose only because both light bulb and window blind have their own IR receiver already installed. In addition, three LEDs with different colors (red, green, and blue) are used. Each color combination is associated to one state of the light bulb and the window blind as shown in
The simulation of the light system adaptation was performed successfully and for each level of the external light the desired states of both the light bulb and the window blind were reached.
The climate system is composed of the heater and cooler where each of them has four states namely: OFF ON Low ON Average ON High
The ambient temperature of the smart room is sensed by a temperature sensor and according to this value the heater and cooler are set to the appropriate configuration as mentioned in Very low: 1T < 0 Low: 0 <= T < 10 Almost low: 10 <= T < 15 Medium: 15 <= T < 20 Almost high: 20 <= T < 25 High: 25 <= T < 35 Very high: 35 < T
The Adaptation task consists of remotely set the both the heater and the cooler to the appropriate state according to the sensed temperature of the smart room and the mapping in
State | Cooler temperature | Heater temperature |
---|---|---|
Off | Off | Off |
On low | 25 | 15 |
On average | 20 | 20 |
On high | 15 | 25 |
Similarly, to the climate and light systems adaptation, the air purifier adaptation system consists of sensing the air quality inside the smart room and setting the air purifier in one of its states as shown in
The interactive subsystem has two components namely the smart TV and the intercom. This subsystem will be controlled using voice commands in adaptation to the physical condition of people with disabilities. As mentioned before in this paper, there will be mainly eight voice commands where seven of them are related to the operation of the smart TV and one for the opening of the door which is closed automatically and do not need a command for its closure.
Commands | Voice words |
---|---|
TV On | T1 |
TV Off | T0 |
Volume+ | V plus |
Volume− | V minus |
Channel+ | C plus |
Channel− | C minus |
Open door | Door |
In this project, we used the Geeetech speech (Module 3) recognition module [
For the simulation of the interactive subsystem, a set of three colored LED (red, green, blue) were used with each combination of lighted led correspond to one of the eight voice command of the interactive system (e.g., the combination (red = on, green = on, blue = off) correspond to the command TV ON and the voice word T1). In addition to be more realistic an infra-red transmitter and an infra-red receiver were used in the circuit to simulate the remote controls of both the smart TV and the intercom.
To the best of our knowledge, the proposed smart room is a complete solution to both health monitoring and comfort improvement for person with stroke living alone. Most proposed solutions focus either on health monitoring or home automation but not both. The proposed system has a modular architecture which enhance its modification, extension, reusability, and maintenance. It encompasses both hardware and software components using IoT technologies with reduced cost which promote its manufacturing on large scale. Some components of the system are ready-made IoT devices especially those used for health monitoring. Adaptation of the reactive system is done according to the current context of the room without explicit intervention from the inhabitant which enhance its intelligence, contributes to the economy of energy consumption, improve the autonomy of disabled inhabitant helps to get rid of the large number of remote-control devices that we use nowadays. The reactive system used a voice-based command method which appropriate for disabled people living alone. The main inconvenient of using such command system is voice interference which is not really present in our case because the disabled person is most the time alone and even though he is surrounded, those present can help by the explicit use of remote controls instead of the voice-based method. In addition, the microphone of the voice-based command system is place close to the disabled person’s mouth to avoid voice interference with TV volume. The health monitoring system is made up of existing IoT devices (mainly smart phone) which promotes continuous health monitoring since all IoT devices should be connected to the internet permanently. The smart watch or health bracelet is also connected permanently to the smart phone and any deviation of biometric values especially blood pressure is immediately noticed and send to patient physician or parents.
IoT is a quantum leap in technological development. One of the most prominent applications of the IoT is smart spaces that are able to interact with inhabitants in order to provide smart services to the user to improve comfort, life quality, energy savings, security and tremendous benefits for the elderly who live alone as well as people with disabilities that are sometimes the result of stroke. Among the valuable uses of IoT is the continuous and real-time health monitoring remotely, especially for people who have a stroke, to avoid having another stroke by following up on blood pressure measurements and some other biological measurements. In this paper, we propose to integrate the advantages of smart spaces and the use of IoT for health monitoring of persons with disabilities due to stroke by developing a smart room which promote the autonomy of disabled people due to stroke by the context-aware dynamic adaptation of devices and equipment. Also, the use of a voice-based command system for interactive devices which is very appropriate for disabled people. These two modules will help to get rid of remote-control devices that require tedious and annoying manual work for people with disabilities, in addition to disposing of a good number of these devices. The third module of the smart room consists of a real-time, permanent health monitoring system based on IoT devices for monitoring among others the blood pressure to avoid the risk of stroke recurrence. The proposed solution has a modular architecture which enhance its extensibility, reusability, and maintenance. As a future improvement to this work, it is possible to add a unit to monitor and track medication doses and remind the times of taking medications.