PRE2016 1 Groep1

Group members

Marissa Damink 0858700 Jochem Meijer 0899769 Nicky Alexander 0886116 Jesse van Kempen 0907453 Jasmijn Kleij 0906848

Work done so far

Week 2 presentatie 1 groep 1

Problem and subject/solution: Our current society is facing the societal dilemma of ageing, given that the ratio of care takers and caregivers is out of balance and will be expanded even more in the future. Considering the fact that technological advancements in health care are taking place, such as better treatment for diseases, also contributes to the increase of elderly people, because their lifespan is extended. Due to this, there will be lack of care givers in the future and therefore it will almost be inevitable that robots are partially taking over tasks. These tasks can not be of great complexity of course, as robots are unable to think autonomous like humans. From this point of view, we have come up with a good use for robots in which the relatively few human caregivers left in the future can focus on more important and risky care tasks which should not be given in the hands of robots, like for instance the use of robots in operating theatres. Our idea will be realized by the use of a ‘Lift Robot’ which will transfer patients, who are not able to relocate themselves anymore, in hospitals and nursing homes by lifting them. The patients can be transferred between for instance a bed and a chair or to places like restrooms etc. However, unlike already existing patient lifts, our ‘Lift Robot’ can work autonomous and will be able to interact with the patient during the transportation by using facial and speech recognition software. Due to this, the care robot is able to react to the patient’s behavior and can adapt or stop it’s action whenever the patient expresses he is uncomfortable or whatever. The human-robot interaction will play an important role in our design cycle. The ‘Lift Robot’ is beneficial for all sorts of groups. The human care givers for example can benefit from it, because they do not have to lift patients themselves anymore which is better for their own health by preventing back pain from heavy lifting and they do not need to assist the transportation anymore which is necessary with the current patient lifts so they can spare precious time and focus on their other care tasks. The health care system itself will also benefit from this technology, because a robot is able to work twenty four seven, it’s indefatigable. Further it does not need a salary to live. Another advantage is that more patients can be helped by human care givers when some parts of their tasks are taken over by robots.

Objective

The main objective of the project is to improve the interaction between a patient and a care robot. To be more specific, it can be divided into three sub-objectives.

The first one is regarding facial expression recognition, which is closely connected to the recognition of certain emotions. We will try to find out what techniques are already used. Also we will go further into the principle of Eckman’s faces, where there are a few basic emotions. If there is not an already existing technology, a new one has to be given thought to. This is important in order to make the robot able to detect facial expressions. Once these expressions or emotions are detected by the care robot, it can react properly to let the patient feel more comfortable. This reaction can vary from a verbal response to a physical action, where the care robot adjusts its lifting mechanism accordingly.

Secondly, it is of great importance to know what the general needs of patients are to feel comfortable as well as the needs of the nurses, so they are able to deliver good and comfortable care. By interviewing these groups of people certain knowledge will be gained, which should also be taken into account by designing the robotic system.

Which leads to the third and final sub-objective, where the optimal design specifications are discussed. On the one hand this includes the mechanical aspect of a comfortable and safe lifting movement and on the other hand the specific look of the care robot; whether it should be more human like and risking that the ‘uncanny valley’ becomes an issue or designing it in a more abstract way.

Use cases

This subject is in many ways related to USE-aspects. First of all there are many different stakeholders and users. The primary users are ofcourse the patients and the nurses. The secondary users are the people who pay for the robot, in this case the hospital management and the government. Also the family members of the patients are secondary users The tertiary users are the maintenance people and cleaners of the hospital. Because they have to take care of the robot, or work around it. The preferences of all these users have to be taken into account. For instance, the patients want the robot to be comforting, while the nurses want it easy to be used. The hospital management wants the robot to be made as cheap as possible and the maintenance people prefer the important components easy to repair or replace. This can lead to conflicts since it can be difficult to take all these preferences into account at the same time. We mainly focus on the preferences of the primary users. The main question is how you let a robot assist them without affecting their dignity and making them uncomfortable. An important aspect is the interaction between the robot and the patient. Do the patients want to be approached by the robot first or do they want to ask the robot for help themselves? Another aspect is how the robot will understand if he’s hurting the patient or not. Maybe he can ask the patient or scan their faces to see if they are in pain. But in this case another question arises: What if the patient is not able to communicate in that way anymore. Maybe the patient had a stroke, which paralyzed a part of his face, and thus the robot cannot read the face anymore. The patient could also has lost the ability to speak clearly to communicate with the robot. These problems should all be taken into account.

Approach

By creating a clear approach, the list of objectives can be answered with back up of different sources. Each method contains a brief description of the task that will be performed.

Interviews One of the objectives is verifiying what kind of needs both patients and care takers have. This will be tested by interviewing a number of nursary homes and private individuals, thus assesing a varierty of problems that can occur, as different institutions can have different sets of problems.The results of these interviews will be taken into account when the design of our ‘’Lift Robot’’ is made.

List of nursary homes around Eindhoven ( https://www.zorgkaartnederland.nl/zorginstelling )

Experiments Furthermore, to test the recognition of emotions, as this is one of the key concepts that will make our ‘’Lift Robot’’ unique, we will conduct an experiment. This experiment makes use of a webcam and software, which is able to spot the micro-expressions of the face of an user, and validate in what state the user is currently. If the outcome of this experiment is postive, thus being a good method to recognize the emotions of the patient, we will use it in our ‘’Lift Robot’’. This makes it possible to tell if the patient is at ease or experiencing a form of discomfort, where the ‘’Lift Robot’’ will react upon.

Companies with software: -> http://emovu.com/e/ -> Software Marissa

Literature study As some of the information of our subject may already been researched, we will conduct a study in literature, by assesing databases on scientific research like the library of the TU/E and Focus. This will give us a better idea about current robots that can perform a similair task and get an idea of their imperfections.

Milestones In the upcoming week this approach will be extended with milestones, setting dates for each of our methods and dividing the tasks so that we will make optimal use of our time.

Conclusion: We think the ‘Lift Robot’ will be an advantage for all kind of groups. Although users can have diverse opinions about the use of robots during the interaction with humans. Some fear there will be less human personal contact, others are more comfortable with the human-robot interaction, because it increases their independence on other humans and streamlines the care process. Our believe is that using this system will make human caregivers be able to work more efficiently. So our goal will be to satisfy as many stakeholders as possible, but the main focus will lie the preferences of the primary users. Of course it should be comfortable for the users by improving the interaction between patient and care robot. To do so we will look at current state of the art by a literature study to complete our design. We will also be taking into account the opinion of the users to make it as comfortable as possible by taking interviews and we will test facial/emotion recognition by experiments. Further, in the upcoming week this approach will be extended with milestones, setting dates for each of our methods and dividing the tasks so that we will make optimal use of our time. We strongly believe that our ‘Lift Robot’ can be usefull for human beings.

NOTES

Meetings

Meeting week 1

Deadlines:

Info (Jochem, Jesse, Jasmijn): Donderdag 14:00 Presentatie Marissa: Vrijdag 12:00 Presentatie Jesse: Zondag

Eerste presentatie: Marissa en Jesse

Iedereen: literatuur onderzoek

Probleemstelling  : Nicky oude mensen, vergrijzing, op langere termijn meer mensen in de zorg nodig dienstverleners in de zorg hebben zware lasten doelgroepen -> verwijzing naar users leuke getallen (hoeveel verzorgers zijn er? Hoe veel hulp is er nodig? Hoe concreet zijn de problemen?) en bronnen noemen nieuwe plan & doel is een robot die meer focust op de interactie door middel van gezichts/emotie herkenning

Onderwerp/ keuze : Nicky mogelijke oplossingen nieuwe plan & doel is een robot die meer focust op de interactie door middel van gezichts/emotie herkenning

Use cases  : Jasmijn primare, secondaire, tertaire users use probleem: mensen die een beroerte hebben gehad zijn moeilijk om gezichten van af te lezen, of mensen die blind zijn kunnen geen oogcontact hebben. Hoe kijken mensen tegen een robot aan? Willen mensen automatisch benaderd worden? Of willen ze eerst toestemming geven?

Objectives  : Jochem · Objective: Describe the specific objectives for the project[1], which should be clear, measurable, realistic and achievable within the duration of the project. Objectives should be consistent with the expected exploitation and impact of the project (see section 2). hoofddoel wat hebben mensen nodig om op hun gemak te voelen? -> interviews herkenning van emoties om mensen op hun gemak te stellen optimale design specifications van een tilrobot/ mechanisme patiënten behoeftes en behoeften van verzorger, en hoe ga je dit implementeren in de robot?

Aanpak  : Jesse interview experiment- webcam en software iets met de robot? planning komt nog, met milestones etc.

Conclusie : Nicky herhaling/ terugkoppeling van probleemstelling: toekomstperspectief

Meeting week 2.1

13-9 Meeting 2

Feedback na de presentatie: De mannen zagen de link niet zo goed tussen robots en gezichtsherkenning. ze wisten niet of facial recognition de interactie echt verbeterd. ze zagen het nut van een til-robot niet zo goed, is dit wel de beste manier? “bed die veranderdt in een stoel” We moeten dingen kwantificeren, meer getallen en bronnen (meer statistieken) Is er vraag naar? Hoe weet je welke interactie wel/ niet goed is.

Updates: Jesse heeft met een verzorgster gepraat, kwamen opvallende dingen uit. Dit zal nog verwerkt worden naar een kwantitatief report. Let op! Altijd als je iemand interviewt moet je een informed consent form laten invullen (staat in de drive)

Nicky heeft met haar moeder gepraat, die twijfelde of de tilrobot wel een goed mechanisme was omdat het gevaarlijk kan zijn. Hoe speelt de robot in op verwondingen?

Het plan: we gaan ons richten op het optimaliseren/ automatiseren van een bestaande passieve tillift. Hierbij willen we een aspect van interactie toevoegen. De precieze interactie moeten we later nog verder uitzoeken, maar een van de dingen waar we gewoon zeker mee aan de slag kunnen is een webcam en eventuele gezichtserkenning software. En een apart scherm waarop de verzorger het in de gaten kunnen houden. Ook nog iets met de afleiding van de patiënt. Aspecten voor verbetering: de verzorger hoeft er niet de hele tijd bij te staan de geruststelling van de patient kan/ moet beter, hoe kan dit? Gezichtsherkenning? Hoe kan het begin/ einde van het tillen beter? Dat de persoon niet de hele tijd heen en weer hoeft te sjorren. Let op! Deze aspecten moeten uit ons onderzoek blijken

Design/ concept: We willen naar een verzorgingstehuis waarbij we zelf opgetild worden door een tillift, en hierbij kunnen we dan onze eigen interactie toepassen. Dit is ter illustratie tijdens onze eindpresentatie

Aanpak: we beginnen met enquetes (quantitatief onderzoek) 1 a 2 interviews over welke aspecten van interactie iets zouden kunnen toevoegen (qualitatief onderzoek) Tegelijkertijd moet er uitgebreid literatuur onderzoek gedaan worden, over de bestaande tillift, wat is goede geautomatiseerde zorg? En iets over de interactie. Wanneer enquetes en interviews zijn gedaan kunnen de statiestieken geanalyseerd worden (mooie cijfers en grafiekjes in eventueel spss) En dan?

Wat we moeten uitzoeken: de goede interactie die we kunnen toepassen hoe wil een persoon behandeld worden, en wat wil de verzorger? Hoe kunnen we de tillift verbeteren zodat de persoon niet helemaal heen en weer hoeft te schuiven? Dat er niet gesjord hoeft te worden!

Planning: Jesse: Voor zaterdag: vragenlijst maken en interview houden Nicky: Voor zaterdag: helpen met de vragenlijst en ook interview houden Vergeet niet de informed consent form Jasmijn: Literatuur onderzoek + artikel over stat-of-the art Jochem: Planning maken, dit wordt donderdag nog besproken: ook in de planning wie wat doet, milestones, deadlines etc Marissa: Voor zaterdag: Vragen voor en enquêtes opstellen en alvast opzet in spss voor de mogelijke verbanden Bellen naar Noldes voor gezichtssoftware

En er moet nog gepresenteerd worden

Literature search

Face Detection and Recogniton (theory) http://www.crcnetbase.com/isbn/9781482226577

Family relations http://ieeexplore.ieee.org/document/6981146/

Reliability facial recognition http://link.springer.com/article/10.1007/s11042-013-1568-8

Experimentation in humanitarian locations: UNHCR and biometric registration of Afghan refugees http://sdi.sagepub.com/content/46/2/144.full.pdf+html

Biometric solution for South Africa's refugees http://www.sciencedirect.com/science/article/pii/S0969476506705859

500,000 refugees are searching for missing family https://www.globalcitizen.org/en/content/ericsson-refugees-search-for-missing-family/

Without Help, Families Face Lonely Search for Europe’s Missing Refugee Children http://thewire.in/42383/without-help-families-face-lonely-search-for-europes-missing-refugee-children/

Facial recognition software spots family resemblance http://www.sciencedirect.com/science/article/pii/S0262407911630236

toepassing: Emotion/face recognition API’s http://nordicapis.com/20-emotion-recognition-apis-that-will-leave-you-impressed-and-concerned/ https://www.microsoft.com/cognitive-services/en-us/emotion-api