Professional Documents
Culture Documents
1
Description of Work
OPEN CALL 2
Application submission starts on:
15 September 2022, 00:00 CET
Submission deadline:
15 November 2022, 17:00 CET
www.hosmartai.eu
Open Call 2: Annex 3.1 – Proposal Template
APRICOT-AI
Page 1
Open Call 2: Annex 3.1 – Proposal Template
Table of Contents
INSTRUCTIONS ............................................................................................ ERROR! BOOKMARK NOT DEFINED.
HOSMARTAI OPEN CALL #2 – EXPERIMENT CALL FOR PILOT ............................................................................ 1
TABLE OF CONTENTS........................................................................................................................................ 2
LIST OF TABLES ................................................................................................................................................ 2
1. OVERVIEW OF THE PROPOSAL (MAXIMUM 1 PAGE) ............................................................................... 3
1.1 EXECUTIVE SUMMARY................................................................................................................................ 3
2. EXCELLENCE/INNOVATION (MAXIMUM 3 PAGES + 1 PAGE FOR IMAGE(S)) ............................................ 4
3. IMPACT & EXPLOITATION (MAXIMUM 2 PAGES) .................................................................................... 8
4. EXPERTISE AND EXCELLENCE OF THE TEAM (MAX. 2 PAGES) ................................................................ 10
5. PROJECT PLANNING AND VALUE FOR MONEY (MAXIMUM 3 PAGES) ................................................... 12
5.1 MAIN ACTIVITIES OF THE PROJECT .............................................................................................................. 12
5.2 VALUE FOR MONEY (MAXIMUM 1 PAGE) ..................................................................................................... 14
APPENDIX A ANNEX: ETHICAL/SECURITY CHECKLIST ............................................................................... 15
List of Tables
TABLE 1: CONSORTIUM COMPOSITION. ............................................................................................................................................... 3
TABLE 3: PROPOSED TEAM ................................................................................................................................................................ 10
TABLE 5: PERSON-MONTH AND PERSONNEL COSTS .......................................................................................................................... 14
TABLE 6: TOTAL BUDGET .................................................................................................................................................................... 14
TABLE 7: ETHICAL ISSUES.................................................................................................................................................................... 15
Page 2
Open Call 2: Annex 3.1 – Proposal Template
To this purpose, Teseo will implement a new system for home care (APRICOT), designed for the specific
needs of the healthcare partner ITIF, building on top of its existing solution for elder care facilities, Kibi,
and integrating it with appropriate home care chronic patient management tools. APRICOT will then
allow caregivers in home care to:
1. Monitor the well-being and habits of care recipients and their trends
2. Provide actionable insights for personalized, proactive assistive care, such as risk of fall and
other markers comparable to Activities of Daily Living evaluation scales, but with the benefit of
being quantifiable, unbiased, and constantly updated
3. Detect falls and alert caregivers in case of emergency
4. Access a nursing record based on an international nursing taxonomy to improve communication
among professionals and with AI systems
5. Receive recommendations meant to simplify the nurse decision making process and foster nursing
plan customization
6. Interact with a vocal virtual assistant to support caregivers and care recipients with reminders
about therapy and personalized screening processes.
ITIF will be involved in the co-creation process of APRICOT and a new operative model to employ it. 50
care recipients and 12 nurses will then use the system in private homes for 4 months. Nurses will visit care
recipients weekly and personalize actions to be taken and frequency of their visits based on APRICOT
data. The most important KPIs are improved QoL, system usability, and reduced caregiver burden and
number of improper hospitalizations. The project is expected to have a major impact on home care
service quality and capacity, and have a strong positive impact on local health systems.
Page 3
Open Call 2: Annex 3.1 – Proposal Template
Families are heavily impacted by the current situation: today in Europe there are 100 million
informal caregivers (20% of the total population). According to a report by Merck, many report
that the care related workload affected negatively their careers (21%), and their finances (30%).
This exceptional pressure is pushing consumers and businesses to look for technological solutions
that can reduce the burden on families, caregivers, and even national health and care systems.
Contrary to popular belief, care recipients are not always against technological solutions,
wearables included. The problem is that without a proper follow-up by familiar and professional
caregivers, their utility is limited. Privacy is a concern, but based on our experience, care recipients
are likely to accept solutions that have a minor impact on privacy if they provide evident benefits
and make them feel cared and safe. Indeed, we already validated that around 70% of our
target population is willing to use an unobtrusive solution and that they would feel more confident
in their daily lives with it.
From the home care service providers point of view, they do everything they can, but their staff
is burdened by time is spent in routine and assessment operations, which as previously mentioned,
can only be periodic, qualitative, and often biased. This lack of visibility has a direct and serious
impact on service quality and capacity, which could be highly improved with data and a new
operative model to make the best use of them. That’s why around 40% of businesses are interested
in a technological aid today, yet only around 6% of them have found something that fits their
needs. Again, traditional Ambient Assisted Living, IoT and wearable solutions offer only limited
help. Bringing AI to home care is the only way to overcome the current issues.
Envisioned scenario
In our opinion, the first step to achieve a better, more sustainable, assistive care for chronic patients
we need a tool unlike anything currently available, able to assess the care recipient autonomy
and general well-being continuously, quantitatively, and in an unbiased manner, and generate
actionable insights for caregivers based on said analysis. At scale, this is only possible using
advanced AI and Behavioral Analysis techniques and integrate them in a new operational model.
Page 4
Open Call 2: Annex 3.1 – Proposal Template
All goals will be made much easier to achieve with the support of three pillars of the HosmartAI
platform, namely:
• the co-creation space, which will allow us to effectively build a user-centred solution, with
robustness, standardization and privacy in mind;
• benchmarking tools, which will allow us to effectively iterate on the first design during
deployment and to fairly assess the results at the end of the pilot,
• the marketplace, which will allow us to effectively spread the excellence of our solution,
and allow us to get early international commercial validation.
The solution perfectly fits into the vision and objectives of the HosmartAI platform:
• It fits perfectly in the assistive care definition, as it has a strong impact on assistance with
activities of daily living (instrumental and not) and health support
• It is a pre-commercial evaluation of the solution and validation of its acceptability and
business viability, with the declared objective to scale it up if successful
• It brings the HosmartAI platform to the real environment of home care
• It aims to deliver to chronic patient care a clear innovation through a rigorous process for
both technology deployment and operational model development.
Technical approach
In practice, the solution we envision is based on the existing Kibi’s AI and behavioral engine, which
analyze data from wearable devices in order to:
1. Monitor the well-being and habits of the care recipient and their trends
2. Provide actionable insights for personalized, proactive assistive care, for example we
provide a risk of fall index, and other markers comparable to Activities of Daily Living
(such as the “drinking activity”), but with the benefit of being quantifiable, unbiased, and
constantly updated
3. Automatically detect falls and alert caregivers in case of emergency
To provide to ITIF a single chronic patient management tool, compliant with the needs of home
care staff, the APRICOT system should license and integrate the Humanavox platform with the
aforementioned AI engine. The Humanavox platform includes:
1. A nursing record based on an international, standard nursing taxonomy to improve
communication among professionals and with AI systems
2. A recommendation system meant to simplify the nurse decision making process and foster
nursing plan customization
Page 5
Open Call 2: Annex 3.1 – Proposal Template
3. A vocal virtual assistant to support caregivers and care recipients with reminders about
therapy and personalized screening processes.
The details of the integration will be defined in the co-creation phase and will be implemented
as much as possible using the HosmartAI technical components at both the Interoperability layer
(Adapter, Interfaces and Standards) and the Data Management and Knowledge Sharing layer
(Hosmart Common semantic data model). As both Kibi’s AI engine and the Humanavox platform
are rather mature, the main barrier will be their integration, which we expect to be made easier
by the tools provided by the HosmartAI platform and the fact that Humanavox platform already
adheres to international standard for the storage of nursing records. The APRICOT system as a
whole is at TRL6 and we expect it to reach at least TRL 8 by the end of the project. The final
system will significantly streamline patient management and provide actionable insight for
customized and preventive care, benefitting chronic patients, businesses operation and caregivers.
There are no age restrictions for the target group. In order to make adequate use of the proposed
technology, preferential criteria will be considered: a fair digital literacy, adequate caregiving
support, an adequate cognitive level.
Other beneficiaries are closely related to the target group:
1. Familiar caregiver: that person who is part of the family or of the friendly network that
stands as a reference point for the person
2. Family doctor: responsible for the diagnostic-therapeutic path, as well as a crucial actor
in the engagement of the patient and in the continuous monitoring
3. Family nurse: health professional specialized in prevention activities and taking charge
of the fragile subject especially in the territorial and home care context
Stakeholder engagement
We will organize questionnaires, interviews, workshops and focus groups in order to define a new
operating model for preventive home care interventions, with the aim of improving capacity,
efficiency and care outcomes. The process will include all stakeholders to identify new problems
that can be solved through the development of ad hoc functionalities. Engagement will take place
during the whole duration of the project and will essentially divided in three phases: (i) the co-
creation phase; (ii) the continuous and iterative improvement phase (during the actual
deployment); and (iii) the final assessment phase.
Gender approach
As you will see in the following pages, the core team is already balanced (3M/2F). Regarding
the team of nurses to be involved, we have the resources to reach full gender balance. As we
implement the actions planned in this project, we will pursuit the following strategy to promote
gender balance at all levels: (i) making a gender equality declaration; (ii) setting goals for gender
balance both in the nursing team and the pool of beneficiaries; (iii) discuss gender balance and
equality at executive meetings; (iv) take into account gender balance in the selection of care
recipient beneficiaries; and (v) ensure that the core team has ownership of the goals.
Page 6
Open Call 2: Annex 3.1 – Proposal Template
Prevent dehydration
Dementia markers
Fig. 1 - Kibi’s Dashboard and a few examples of how behavioral data can be correlated to common
conditions in order to support preventive care
Fig. 2 – The home care patient management platform with which we would like to integrate Kibi
Page 7
Open Call 2: Annex 3.1 – Proposal Template
Expected benefits
The fragile person, understood as a person suffering from chronic disease, with a complex
therapeutic regime or at risk of developing chronicity, needs continuous assistance over time to: i)
promote a dignified quality of life, ii) guarantee health well-being, fundamental to independent
living, iii) being able to continue living at home as long as possible in full autonomy. Therefore,
the aim of the project is to strengthen the territorial network taking charge of the person's health,
drastically increasing preventive home care activities through minimally invasive actions with a
strong impact on daily habits and well being. The use of AI technology allows this model to be
fully scalable and favors the reuse of collected data, which could be of great importance in all
health domains. The periodic presence of a health worker guarantees that actionable on care
recipients needs, often still experienced as latent or of negligible importance, are followed up by
adequate actions according to a preventive care model.
Impact
Short-term
Measurable outcomes
• Solution acceptability by care recipients
• Solution liking and usability by all stakeholders
• Liking of the care service as whole as transformed by the applications of our solution
• Measurable reduction in the familiar caregiver burden
• Measurable increase in the care recipient Quality of Life
• Measurable increase in awareness of the degree of autonomy and well-being of the care
recipient, by both himself and his family
• Measurable decrease in the propension of the family to hospitalize the care recipient
• Measurable reduction in the number of improper hospitalizations compared to the data
of the 6 months preceding the pilots
• Increased number of early diagnosis by partner family doctors compared to a test group
Exploitation potential
This project is born with the purpose of validating a new hybrid care model, strongly accelerating
the synergy between technology and health professionals. The service developed in this pilot aims
to be a scalable and replicable model that empowers the healthcare team with adequate tools
for non-invasive long-term preventive home care, improving care outcomes and efficiency, through
Page 8
Open Call 2: Annex 3.1 – Proposal Template
flexibility and a relative customization of at-home interventions according to specific needs. Such
model has great economic value, not only for project partners, but for the HosmartAI platform
and the wider home care market, and could easily be scaled internationally with eventual minimal
adaptations based on the peculiarities of each country home care market and regulations.
In the short-term, project partners will benefit directly from the project: Teseo will open up a new
market by validating a killer use case of its technology, while IFIT will benefit from an ad-hoc
solution, allowing them to make their operations more efficient and improve their service quality,
benefitting also customer retention and service differentiation in a largely commoditized market.
As a consortium, project partners are obviously excited by the possibility to scale up the solution,
directly, though the direct offering of a hybrid technological home care service; and indirectly,
by offering to third party home care providers the tools developed in the project as SaaS solution.
The HosmartAI platform plays an important role here thanks to its marketplace and because it
favors data exchange, largely increasing the potential for synergies with other healthcare
institutions and tech providers which could be interested in the kind of data we generate. All things
considered, by continuing to work on the solution even after the deployment phase of the project,
we believe we will be able to begin commercial exploitation in Italy right after the end of the
project without any further funding. International scale-up will require private investment but Teseo
has a strong track record in this regard and the validation and traction obtained in the pilot will
make the process much easier.
As a consortium, we also believe that the results of the pilot will be a very powerful marketing
tool. As such, it is our intention to make publicly available the operative model we will develop
and the results achieved. This information will be published through the usual mix published articles,
online presence, and social media marketing.
As shown before, technically our solution is fundamentally different from traditional AAL, IoT and
wearable solutions, which are increasingly seen as commodities by the market. By bundling
APRICOT with a novel operative model, we are further differentiating from alternative solutions,
positioning us as the ideal one for home care, and reducing barriers to entry for SMEs with little
capacity to build the model themselves.
• What are the Key Performance Indicators you will use to measure your project’s
success?
Page 9
Open Call 2: Annex 3.1 – Proposal Template
Teseo is an innovative SME founded in 2015 as an official spin-off of the University of Genoa, whose
mission is the development of digital solutions to support caregivers. The company specializes in Artificial
Intelligence, Machine Learning and IoT technologies, aimed at behavioral analysis and Ambient Assisted
Living, as evidenced collectively by more than 30 publications on the subject (more than 100 in total)
and 2 international patents. The company has always been committed to two main business lines: the
development of its own solutions in the healthtech field and technological consulting services in the fields
in which it is an expert. Its main product is Kibi, a digital platform that enables caregivers to provide
personalized and high-impact care, alerting them in case of emergency and providing actionable insights
about their habits and needs. The Kibi project was awarded in 2019 a public grant for Silver Economy
projects by the municipality of Genoa, and the system has already been successfully adopted by
selected partners. Other significant projects are:
• Research partnerships with important partners such as the Azienda Ospedaliera Ospedali
Galliera, a national excellence center in geriatrics, and the Casa di Cura Privata del Policlinico
di Milano, a leading rehabilitation center.
• Successful actions in the scope of European grants and tenders, such as the Near2U pilot in the
scope of the Interreg I-Care-Smart project, and an ongoing co-creation and pilot deployment
project funded by ZIP PON-METRO, in collaboration with the Fondazione Chiossone, to support
with its technology patients affected by visual impairments.
Team members relevant experiences:
• Antonella Giuni. She has almost 40 years of experience in R&D and Strategic Marketing.
During her career she worked on a dozen large projects in MRI, Robotics and Ambient
intelligence. These include the development at Ansaldo Elettronica Biomedicale of the first MRI
system produced in Italy and the launch of the first MRI system dedicated to the extremities at
Esaote.
• Alessandro Sperindé. He is an expert in design and software development for embedded
systems. He has more than 15 years of experience divided between academia and industry. As
a researcher he has participated in relevant roles in large projects often in association with large
private companies (H2020, Thalès Alenia Space, Leonardo, NATO S&T CMRE).
Page 10
Open Call 2: Annex 3.1 – Proposal Template
• Mohamad Alameh. He is a PhD with expertise in AI-on-chip, particularly in the field of artificial
touch. He participated as a researcher in the study of the fine manipulations of the human hand,
developing an advanced sensor solution based on wearable inertial sensors.
• Alessio Capitanelli. Former researcher at the University of Genoa in the field of artificial
intelligence and robotics. At Teseo since 2018, he has since taken the role of Product manager
of Kibi, being the interface between the technical team, the sales team, and the various
stakeholders. He has international experience organizing focus group and interviews, both for
Teseo and in the context of several funded projects.
Main publications
Bruno, B., Mastrogiovanni, F., Sgorbissa, A., Vernazza, T., & Zaccaria, R. (2013, May). Analysis of human
behavior recognition algorithms based on acceleration data. In 2013 IEEE International Conference on
Robotics and Automation (pp. 1602-1607). IEEE.
Mastrogiovanni, F., Sgorbissa, A., & Zaccaria, R. (2010). A cognitive model for recognizing human
behaviours in smart homes. annals of telecommunications-annales des télécommunications, 65(9), 523-538.
Buoncompagni, L., Capitanelli, A., Cristofanini, M., Giuni, A., Mastrogiovanni, F., Motolese, C., ... &
Zaccaria, R. (2020, November). Persuasive and Polite Sentences to Drive Human-Robot Interaction in
Smart Homes for Elderly Care. In AIRO@ AI* IA (pp. 33-38).
Buoncompagni, L., Kareem, S. Y., & Mastrogiovanni, F. (2021). Human Activity Recognition Models in
Ontology Networks. IEEE Transactions on Cybernetics.
Bruno, B., Mastrogiovanni, F., & Sgorbissa, A. (2014). Recognition of human activities through wearable
accelerometers. Robot and Human Interactive Communication.
Il Tuo Infermiere di Famiglia is a Cooperative of Nurses that operates mainly in the Lombardy region,
providing health services on a continuous and single performance basis, which is interested in adopting
new technologies. The cooperative is a pioneer of telemedicine which integrates into its daily operations
to carry out rapid and timely diagnostics remotely, directing patients to the most appropriate care and
the most suitable therapeutic path. Delia Ragni is the current president of the cooperative and a certified
nurse with 10 years of experience, both as a practitioner and in operations organization.
As a single team, Teseo and ITIF possess all the required expertise, technological and in the field of home
care operations to successfully implement the project and deliver timely and quality results. In particular,
Teseo not only will provide the technology, but has the expertise and experience to integrate various
systems as required by ITIF, set up the required infrastructure, provide technical support and train end
users as needed by the project. On the other side, ITIF provides a large pool of domain expertise on
home care and healthcare in general, deep understanding of the needs of the target population, ability
to interface fruitfully with all the stakeholders (including care recipients, their families and family doctors),
and a large pool of customers which can be recruited for the pilot. Teseo itself is no stranger to field of
healthcare and has ongoing relationships with several stakeholders, in Italy and abroad (mainly
Switzerland and US), which will be essential to the scale up of the solution after the project end.
The core team involved in the project will be structured as a single cross-functional unit, and the
development of both the APRICOT system and the new operational model will follow state-of-the-art
practices of agile iterative product development.
Page 11
Open Call 2: Annex 3.1 – Proposal Template
Deliverable: Technical report of the APRICOT system v1; Report of the co-creation activities and
proposed operative model for the pilot.
Page 12
Open Call 2: Annex 3.1 – Proposal Template
the duration of the project in accordance with the evolution of the operating model. ITIF coordinator will
organize the training session with the nurses and assess their effectiveness.
Deliverable: Publications (1 newspaper + 1 online), APRICOT Landing page, Social media posts
Page 13
Open Call 2: Annex 3.1 – Proposal Template
* 12 Nurses will be involved in the project. For sake of brevity and budget clarity we report them as a
single expenditure. As the total PM effort is 4, each nurse will be expected to be involved in the project
for 1/3 PM. We are available to provide the full list of persons to be involved as nurses upon request.
Page 14
Open Call 2: Annex 3.1 – Proposal Template
Page 15
Open Call 2: Annex 3.1 – Proposal Template
Ethics
If you have entered any ethics issues in the ethical issue table, once selected for funding you
will need to have ready to submit:
If these documents are not in English, you must also submit an English summary of them
(containing, if available, the conclusions of the committee or authority concerned).
If you plan to request these documents specifically for the project you are proposing, your
request must contain
Security
Page 16