Robots in Healthcare – A Technological Review (sample)

Introduction

As of today, hospitals around the world are experiencing a worker shortage leading to a high burden on medical departments and potentially harming the quality of care. In addition, we are observing consistent demographic growth in the elderly population in Israel based on the Israeli Central Bureau of Statistics (CBS). This trend is projected to continue over the next few years as can be seen in the adjacent graph. In order to deal with this matter, several technologies on the market aim to improve certain aspects of healthcare services. A possible solution is integrating logistic robots in hospital departments in order to reduce the workload among healthcare workers. Since the start of the COVID-19 pandemic, the field of Telemedicine (remote diagnosis and treatment using telecommunications technology) has developed vastly to reduce infection rates and allow for convenient service. In a 2022 survey from the US, patient and caregiver satisfaction using telemedicine was examined and showed positive results. Further information can be found in Appendix. In recent years, nursing robots gained popularity thanks to their ability to improve medical treatment, reduce workload and reduce the risk of infectious diseases (such as COVID) using remote medicine This report will review robots used in healthcare service, the market for these robots, existing solutions, and possible use scenarios. This report aims to identify a robot with several key abilities including autonomous movement, obstacle detection, transporting equipment such as medications, identifying patients, telemedicine – video calls with medical staff, monitoring and tracking patients, measuring vital signs (such as temperature), and more. 

Literature review

As part of the research, a literature review was conducted focusing on articles, research, and reviews that could indicate and highlight the benefits of using healthcare robots, with the following findings:

Reducing the exposure of staff to pathogens

  • The use of robots and telemedicine, in theory, can reduce the risk and exposure of workers to infectious disease. We found evidence that the use of robots helped to prevent the spread of pathogens. For example, the robot Lio (further reviewed later in this report) helps to disinfect surfaces using UV light and measures the temperature of patients\visitors. Other robots perform deliveries inside the hospital and by doing so, reduce the exposure of the staff to covid


Usage of robots during the Covid-19 pandemic

  •  A review published in 2022 in the “Journal of Rehabilitation and Assistive Technologies Engineering”, examined the advantages and disadvantages of integrating robots among the elderly population before and during the
    pandemic. According to the review, Telepresence could benefit patients by creating closer interaction between patients and caregivers\families and alleviating loneliness while maintaining isolation protocols. The review found that telepresence had a greater effect over phone calls or other technological means. Thanks to these robots, families had a better understanding of the medical process. A 2022 research article published in “JMIR Human Factors” examined the level of satisfaction of patients who used telemedicine during the pandemic. Out of 368 participants, 35% were satisfied and 47% were very satisfied. 77% reported that they “look forward to using telehealth in the future”. A higher satisfaction rate was shown among younger participants, a population that is known to adopt new technologies faster than older participants.
  • A review published in 2022 in the “Journal of Medical Internet Research” examined the use of robots in clinical care, the review included 33 different studies. The robots were shown to reduce the team’s response time, allowing earlier intervention, higher patient survivability, and lower mortality rates, especially during night shifts, when there is a staff shortage. The review describes other uses, such as telemedicine and strengthening the bond between isolated patients and their families. The participants (caregivers and patients) stated high satisfaction and mentioned that they would use telemedicine in the future. Additional uses included providing psychological therapy for patients in isolation. 

Assisting nursing staf

  • Hospitals are required to provide service 24 hours a day, even during difficult times (such as the covid pandemic), and are required to provide optimal treatment for each patient. Robots could help hospitals deal with such challenges.
  • A financial analysis published in 2015 in “Critical Care” evaluated the economic impact of automated-drug dispensing systems (ADS). Thanks to these robots, 14.7 hours were saved per day/33 beds. After 5 years a total of €510,404 had been saved.
  • An article published in 2017 in “Clinics” demonstrated the use of a similar robot in a hospital which saved $14,444 a year and saved 4 hours of work per day. A significant portion of the tasks in hospitals is logistics (transporting equipment, specimens, meals, medicines, etc), some of them are repetitive and performed today by hospital staff members. Using robots could make the process more efficient, reduce workload and even reduce financial costs.
  • A research article published in 2021 in the “Journal of Nursing Management” followed 733 nurses and concluded that 33% of their time was spent on simple logistic tasks such as transporting specimens. Thanks to recent developments, robots can work together with the nurses and perform such tasks; for example, the robot MOXI – a robot equipped with a robotic arm (further reviewed later in this report). 
  • A research article published in 2015 in Studies in “Health Technology and Informatics” compared automated and manual vital sign measuring (blood pressure, pulse, temperature, weight, and oxygen saturation) in hospital wards. The study found that automated vital signs collection was faster (158.7±67.0) than a manual collection (4079.8±7091.8 s). Furthermore, the group that used manual collection had a 30% error rate while entering the data in the system, which did not occur in the automated team.
  • A study published in 2014 in the “Journal of Evaluation in Clinical Practice” showed that administering medications using an automated robot could reduce the error rate by 53% compared to manual administration.
  • The company Aethon, designed a logistic robot for hospitals named TUG (further reviewed later in this report). Aethon shared the results of integrating TUG in hospital wards, such as decreased drug delivery time from 74 to 30 minutes and saving 6,123 work hours for the nurses during its first year.

The demographic changes previously described, and the reviewed benefits above, highlight some of the challenges hospitals face in the near future. Challenges that can be met with the use of robots. Using robots we can reduce staff exposure to hazards and infectious diseases, improve patient satisfaction, reduce workloads, and optimize certain processes.

 

 

Business review

Market Research and Future Forecast According to market research by “Research and Markets” published in 2022, as of 2019 the market for robots in healthcare is valued at $482.939 million and is expected to grow to $2,232.717 million by 2026 with a CAGR (compounded annual growth rate) of 24.45%. The covid pandemic had a positive impact on the market which led to a growth increase. The pandemic strengthened the market and highlighted the need for robots in healthcare. Furthermore, we are witnessing additional investments in private companies in the field, such as “Diligent Robotics”, which designed MOXI. Last April, the company raised 30 million dollars as part of series B. “Temi”, the company behind robot Temi (further reviewed later in this report), raised 15 million dollars as part of series C.

An article published in 2017 by the “American Journal of Medical Quality” presented a forecast for nurse shortage in the USA by 2030. The article is based on previous forecasts and newer models. A shortage of half a million nurses is expected by 2030 in the USA alone with a possible 37 states suffering a severe shortage. A report by the Israeli Ministry of Health from 2022 reviewed the long-term planning of human resources in health professions. According to the report, the number of nurses per 1,000 individuals grew in recent years (5.6 per 1,000 individuals), yet is still lower than the ratio in OECD countries (9 per 1,000 individuals) as of 2020. Nursing and logistic robots could perform some of their tasks and aid medical staff in dealing with the nurse shortage by reducing the work burden. As of 2020, adults above the age of 65 were 9% of the global population. Based on OECD data, over time the elderly population takes up a greater portion of the population in developed countries. In Israel, they represent 12.06% of the population, in the USA 16.89%, and in Europe 20.69%. The UN estimates that by 2050 their portion of the population will rise to 16%. According to the Israeli CBS, in 2015, adults above 65 years old were 11.1% of Israel’s population and this number will grow to 14.3 and 15.3 in 2040 and 2065 respectively.

This increasing growth would require additional nursing staff for hospitals and clinics, to support the growing elderly population. Global population growth is a key factor for the higher demand for healthcare robots and the market’s growth.

Market research states major players in the field: Diligent Robotics, Toyota Motor Corporation, RIKEN-SRK, SoftBank Robotics, Panasonic, Fraunhofer IPA, and Aethon. Some of the companies and their products will be reviewed in this report.

Hospitalization market and nursing facilities market potential.

Hospitals, HMOs (health maintenance organizations), retirement homes, and other facilities, might take interest in healthcare robots. In order to understand the investigated market size, we based our estimation on the number of beds each facility has, which can indicate the number of potential users. Our results are based on a report by the Israeli Ministry of Health for the year 2022, and another report, written by the National Headquarters Against Corona in the year 2020: Health systems around the world have similar potential and might be even greater. Appendix 3 includes data by the OECD that present the number of hospitalization beds and beds in retirement homes per 1,000 inhabitants among OECD countries.

Robots in Healthcare – Technological Review

Telemedicine Robots

4.01 Temi

Background: Temi is a robot that can move autonomously, engage remote video calls between patients and staff members in covid wards, transport equipment and entertain patients while they wait for their appointment. It is possible to install additional systems on the robot and the company highlights several examples on its website such as a hand sanitizer device, a thermal camera to measure body temperature, a room disinfection system, and more.

The robot is designed by the company Temi and imported into Israel by One Robotix.

Places in use: The product is in use in Ichilov hospital in the pediatric Ophthalmology department (entertaining patients), the retirement home Amal (telepresence, occupational therapy), Carmel hospital (directing visitors) and in the future in Sheba hospital (directing visitors).

Additional information about the robot can be found in a demonstration video and in the company’s website. A different video presents the use of Temi in Korea to assist and maintain covid restrictions (measuring body temperature, remote diagnosis using video calls, and more).

Key features:

  1. Movement – The robot moves autonomously, equipped with a number of depth cameras and a LiDAR sensor that allows for free movement in space while avoiding obstacles, without the need for an operator.
  2.  Carrying cargo – Capable of carrying cargo up to 3 kg. The cargo is placed on a shelf behind the monitor. It requires an operator to load and unload the cargo. The robot can transport the cargo to its destination, where a worker can unload it.
  3. Telemedicine – Capable of engaging remote video calls.
  4. Identifying patients – The robot can identify patients based on face detection using depth cameras. It should be mentioned that this is not an ideal solution, since the identification process is based on the relative position of the patient to the camera, the patient needs to look directly at the camera, lighting in the room might affect this feature, and identifying patients who use mask\are intubated might be a challenge. In addition, prior to identification, it’s required to enter 3 pictures of the patient into the system. A possible solution could be scanning a QR code – each patient would receive a QR code printed on his\hers bracelet and the robot will scan the code for identification. QR scanning could help in some scenarios, yet the patient still needs to present the QR code to the camera. In case the process
    fails, an operator can initiate a video call to identify the patient. To verify this solution, we contacted the manufacturer who confirms it is possible to scan the QR code, yet it might require additional development. Another possible solution is a radio frequency identification device – RFID. Each patient will receive a bracelet that the robot can scan and identify the patient
  5. Language – The robot supports Hebrew and English and is based on a dialog between the user and the system (similar to voice assistants such as Alexa, Siri, etc). The robot supports Arabic and Russian languages (additional development might be required to support all functions).
  6. Physiological signs – the system does not contain built-in solutions to measure physiological signs. It is possible to extend the system’s abilities by connecting external devices via USB\BT\WiFi. It is worth mentioning that the robot is limited by the dimensions and weight of the desired device. Lightweight systems can be installed, the company presents several examples on its website, such as a CIO2 Disinfection (top picture), a thermal camera (bottom picture), a room disinfection system, and more. Additional examples can be found on the company’s website.
  7. Routines – Operators can plan routes and routines for the robot to automatically execute. For example, a night routine in which the robot patrols the ward to make sure none of the patients have fallen from their beds.
  8. Connecting to external systems – Chameleon, Paradigma, and any device that connects via USB\BT\WiFi.
  9. Battery life – up to 8 hours without charging.
  10. Entertaining patients – Temi can interact with patients and entertain them while they wait for their appointment, answer questions, play videos, etc.
  11. Network – WiFi\SIM card.
 

Price: $5,000 per device and an additional monthly fee of $100 for software licensing.

Possible scenarios for the use of telemedicine robots:

  1. Communicating with and identifying patients – a patient could be able to click a button and call the robot, which will arrive at the patient’s bed and identify the patient. Afterward, the robot will gather initial data about the patient’s complaint, based on a predefined dialog algorithm and the patient’s answers. Clinical staff will receive the information and reduce the data collection time. In case the information is not clear, an operator could initiate a video call and converse with the patient.
    Issues that arise from the scenario: the limited ability to identify patients – since the process is based on face recognition, the process depends on different criteria such as the relative position of the patient to the camera, the patient needs to look directly at the camera, lighting in the room might affect this feature, and identifying patients who use a mask\are intubated might be a challenge. Furthermore, it requires the creation of a dialog algorithm between the patient and the robot. It requires the creation of an interface between the call button and the robot’s system.
  2. Protection from infectious disease – staff entrance to covid ward requires a lot of protection beforehand. In certain cases, such as giving medicines to a patient, the staff could use a robot. The robot would carry the medicine to the patient’s bed, and a staff member would identify the patient using a video call and instruct him to take his medication. This could reduce exposure to infectious diseases, such as Covid-19, and protect the medical staff.Issues that arise from the scenario: The robots reviewed in this chapter are unable to open doors and require staff members to help them\modify the ward for the robot needs. As described above, patient identification is limited. The robots Pepper and REEMAN can communicate with patients, but unlike Temi, they cannot carry cargo.
  3. Night patrol – It is possible to configure routines, such as patrolling to assure none of the patients have fallen from their beds, or to assure there isn’t unusual activity in the ward. The robot can patrol during the day and ask patients about their well-being. The information will be sent to the staff in real-time, allowing them to examine the information and respond. Issues that arise from the scenario: Theoretically the robot is capable of doing such tasks, yet while integrating the system the staff should define what is an unusual activity (for example, loud and unusual noises) and how the robot can identify it. We should bear in mind that the robot’s abilities are limited by the sensors it has. The robot is capable of certain tasks such as room mapping and using a microphone and camera.
  4. Reception and Directing Patients – The robot could greet visitors at their arrival at the clinic\hospital and direct them to their destination. While waiting in line, the robot could answer questions (using information preset information), play videos, and make their waiting more pleasant. Issues that arise from the scenario: We believe that older people might have a harder time using the robot in an effective way. A Qualitative Study published in 2021 in “JMIR mHealth and uHealth” explored how adults use voice assistants (such as Alexa, Siri, and Google home) for the first time. The study included 18 participants above the age of 74. The overall first response to a voice assistant was positive, thanks to the simplicity of a speech-based interaction, yet they had difficulty using it due to their lack of understanding of its technological limitations. An Article published in 2020 describes the robot Lio – a logistic robot equipped with a robotic arm for hospital and home use (further information can be found in the full report). The article presents a challenge to older people to use and operate the robot because of the robot’s limited vocabulary and the manner in which users may address him. Lio does not proactively approach people for interaction. Because of this, a reduction of interest in the robot was observed among some people as time passed. 

Logistic\Nursing Robots.

4.06 MOXI

Background: MOXI is an autonomous robot designed to assist nursing staff. The robot is equipped with a robotic arm, allowing him to pick up objects, open doors, and transport lab specimens – without an operator. In addition, the robot can execute preset routines. According to Diligent Robotics, the company behind MOXI, 30% of nursing staff’s time is spent on logistic tasks, part of which can be done using MOXI.

Places in use:

  • Texas Health Dallas.
  • The University of Texas Medical Branch (UTMB Health).
  • Houston Methodist Hospital.
  • In May 2022 Bloomberg published an article describing “ChristianaCare”, a facility that received a $1.5 million grant from the American Nurses Foundation to purchase MOXI robots and staff training. Last February, ChristianaCare acquired 2 robots and now they acquire 3 additional robots. In the article, Katherine Collard, chief nursing informatics officer at ChristianaCare anticipates that the robots will complete up to 200 delivery tasks a day. We can see the grant and the additional robots acquired as a vote of confidence by the American Nurses Foundation in the robots and in their contribution to the facility.
 

additional information about the robot can be found in a demonstration video and in the company’s website.

The key finding from studies describing the usage of MOXI:
A case study published by the “American Nurses Association” in 2020 reviews the influence of robots in the field of nursing. The case study addresses experiments done with MOXI. According to the document, MOXI can transport cargo such as boxes, water bottles, linen, deliver and gather specimens (urine, stool. sputum), a blood draw kit, a feeding-tube change kit, and wound care supplies.

The case study describes part of MOXI’s tasks in hospitals:

Restocking inventory – each day the MOXI checked the amount of each item in the ward’s inventory and by 7:30 MOXI sent an email to Materials Management requesting the necessary items to bring the unit up to minimum levels for these
items. After an hour, Moxi traveled down to Materials Management to collect the supplies and deliver them back to the supply room on the unit.

Patients fall risk – Every 30 minutes MOXI received a list that included each patient’s most recently documented fall risk status, Moxi delivered fall precaution supplies (slippers and fall risk bracelet) for each patient whose status changed to moderate or high. After the fall risk status changed, MOXI visually scanned the fall risk sign outside the patient’s door and verified whether the sign matched the patient’s most recently updated fall risk status.

Part of these tasks are shown in the pictures below: 

 

The case study does not contain a funding statement, but it mentions that Diligent Robotics cooperated with different hospitals. The level of involvement is unclear.

Key features:

  1.  Movement – The robot moves autonomously, equipped with a number of depth cameras and a LiDAR sensor that allows for free movement in space while avoiding obstacles, without the need for an operator.
  2. Carrying cargo – MOXI is designed for logistics tasks and transporting cargo as described before. The robot has a robotic arm, allowing it to transport supplies without an operator.
  3.  An important note according to the following article: the robot requires a dedicated room for efficient operation – the room should contain cataloged and marked supplies that MOXI can access and identify independently.
  4.  Languages – English.
  5. Physiological signs – The robot does not have the ability to measure physiological signs.
  6. Routines – Operators can plan routes and routines for the robot to automatically execute.
  7. Network – WiFi.
 

Possible scenarios for logistics\nursing robots:

  1. Logistic support – the robots can transport lab specimens\cargo and shorten the time workers spend on logistic tasks. Operators can set up a schedule or routine and send the robot to a specific location\patient. The
    robot moves autonomously and can transport medications or specimens to patients. Based on an article published in Bloomberg, MOXI can carry up to 31 kg and work 22 hours without charging. Based on information by Diligent Robotics, MOXI can perform different logistic tasks, some of which require modifications to fit client needs. The robot TUG has a high carrying capacity (up to 635 kg) and is able to transport supplies around the ward. The robot MOXI can have routines programmed. Supplies are protected in a drawer with a biometric lock, allowing only staff members access. The robots Lio and Care-O-Bot 4 can carry lightweight using their robotic arms. Issues that arise from the scenario: MOXI requires a dedicated room for efficient operation – the room should contain cataloged and marked supplies that MOXI can access and identify independently. It is possible that Lio and Care-O-Bot 4 require a similar room. The robot Lio and Care-O-Bot 4 use facial recognition to identify patients, therefore they suffer from the same limitations described in the chapter on telemedicine robots. TUG requires a staff member for receiving and loading supplies. Additionally, it cannot identify patients.
  2. Protection from infectious disease – staff entrance to covid ward requires
    a lot of protection beforehand. In certain cases, such as giving medicines to a
    patient, the staff could use a robot. The robot would carry medicine to the patient’s bed (some of these robots can open doors using their robotic arm, if necessary), and the robot will identify the patient and instruct him to take his medication. This could reduce exposure to infectious diseases, such as covid-19, and protect the medical staff. Issues that arise from the scenario: The robot Lio and Care-O-Bot 4 use facial recognition to identify patients, therefore they suffer from the same limitations described in the chapter on telemedicine robots. TUG cannot identify patients or open doors.
  3. Physiological monitoring – customers can install external devices on the robots, and add new functions and abilities. For example, a thermal camera to measure body temperature that connects directly to the robot and transfers the data to a staff member. Issues that arise from the scenario: most of the companies we contacted for this project, state that this should be possible, yet it also depends on the type of the desired device. Each sensor\device might require different integration for communication, power supply, weight, position on the robot, etc. 4. Night patrol – It is possible to configure routines, such as patrolling to assure none of the patients have fallen from their beds, or to assure there isn’t unusual activity in the ward. The robot can patrol during the day and ask patients about their well-being. The information will be sent to the staff in real-time, allowing them to examine the information and respond. Issues that arise from the scenario: Theoretically the robot is capable of doing such tasks, yet, while integrating the system the staff should define what is an unusual activity (for example, loud and unusual noises) and how the robot can identify it. We should bear in mind that the robot’s abilities are limited by the sensors it has. The robot is capable of certain tasks such as room mapping and using a microphone and camera. The robot TUG cannot patrol the hospital ward.

Summary and conclusions

  • The report began with a literature review, in order to examine, and assess the possible advantages of integrating robots in hospitals. In our research, we found that healthcare robots contribute in different fields such as telemedicine, reducing exposure to infectious diseases, and saving time – as we found in recent studies. Additionally, the data projects population growth and specifically, the growth of the elderly population around the world and in Israel. This information and the shortage of nurses highlight the need for new solutions to deal with these factors.
  • As part of the report, we introduced technologies that include several leading robots in the fields of telemedicine, nursing, and logistics. The report provides specific information about each robot and its advantages, including findings from articles, research, online information, and information gathered by proactively contacting the different companies, to have a better understanding of the robots and their abilities. Finally, we presented different use scenarios along with the limitations that arise from these scenarios.
  • As part of reviewing technological solutions, we examined a wide range of robots in different development stages. Between these solutions, the most
    mature products we found are Temi, MOXI, and TUG. MOXI and TUG are the most prevalent among literature papers and online articles.
  • Most of the robots specialize in one field: logistics or telemedicine, and usually their abilities in the other field are limited, for example, MOXI is a robot with high logistic abilities, yet it lacks telemedicine communication with patients and nursing abilities. On the other hand, Temi focuses mainly on interactions with patients and is supported by different languages, but its logistics abilities are significantly fewer than those of MOXI.
  • Telemedicine is a common feature, yet it is based on a basic process of patient identification (with pre-existing pictures and a camera). It might be necessary to develop a different solution for patient identification to solve the current issues.
  • The robots’ ability to measure physiological signs is limited, and mostly ends up in measuring body temperature, it might be possible to expand this ability using external systems on the robot.
  • The vast majority of manufacturers support a modification process, allowing the customers to extend the robot’s abilities and perform additional tasks such as room disinfection, patrols, and more.
  • Based on recent information, the healthcare robots market is headed toward logistic applications. Communicating with patients is not a field under focus based on the results of this study.
  • Advantages and distinctive features of Temi compared to other solutions -Currently, Temi is in use by several hospitals and health centers in Israel and has gone through adaptation to the Hebrew language. The robot answers most of the demands and features (lightweight transportation, remote video calls, entertaining patients, patrol, measuring temperature). The main limitation lies in its ability to carry cargo and external devices, due to the small dimensions and weight limitation. The price of the system is a key advantage, Temi had the lowest cost among the robots we reviewed.
  • In light of our findings, we believe that Temi is the optimal solution for telemedicine, lightweight transportation, autonomous operation, and price. However, Temi has low logistic abilities and it might be worth examining other logistic solutions such as MOXI or TUG, which represent more advanced logistic abilities.
  • However, the robot Gary is a unique example of a robot that combines telemedicine and logistic abilities. Gary could gather physiological information (temperature and oxygen saturation), yet the product is still under development and according to the company, these features may not be available in the final product. Gary could provide a suitable solution, yet the price tag, which is significantly higher than Temi, and the fact it’s still under development, are matters to consider.

References

  1. https://www.gov.il/he/Departments/General/manpower-forecast-2021
  2.  https://www.cbs.gov.il/he/mediarelease/DocLib/2017/138/01_17_138b.pdf
  3. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8941569/
  4. https://www.mdpi.com/2227-7080/9/1/8/pdf
  5. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8043496/
  6. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7312924/#
  7. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8027549/
  8. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4563942/
  9. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7442501/
  10. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5666447/
  11.  https://onlinelibrary.wiley.com/doi/10.1111/jonm.13451
  12. https://pubmed.ncbi.nlm.nih.gov/23823371/
  13.  https://pubmed.ncbi.nlm.nih.gov/27637980/
  14.  https://pubmed.ncbi.nlm.nih.gov/24917185/
  15. https://www.aethon.com/wp-content/uploads/2014/08/UMMC-Case-Study-2011b.pdf
  16. https://www.researchandmarkets.com/reports/5547597/nursing-robots-market-forecasts-from-2021-to?utm_source=GNOM&utm_medium=PressRelease&utm_code=9zcsqt&utm_campaign=1666290+-+Global+Nursing+Robots+Market+(2021+to+2026)+-+Featuring+Diligent+Robotics%2c+Softbank+Robotics+and+Aethon+Among+Others&utm_exec=jamu273prd
  17. https://www.grandviewresearch.com/industry-analysis/robotic-nurse-assistant-market
  18.  https://pubmed.ncbi.nlm.nih.gov/29183169/
  19. http://www.israelhpr.org.il/wp-content/uploads/2022/02/%D7%A1%D7%99%D7%9B%D7%95%D7%9D-%D7%94%D7%9E%D7%9C%D7%A6%D7%95%D7%AA-%D7%94%D7%A6%D7%95%D7%95%D7%AA%D7%99%D7%9D-%D7%9B%D7%95%D7%97-%D7%90%D7%93%D7%9D-%D7%91%D7%9E%D7%A7%D7%A6%D7%95%D7%A2%D7%95%D7%AA-%D7%94%D7%91%D7%A8%D7%99%D7%90%D7%95%D7%AA.pdf
  20. https://data.oecd.org/chart/6Jwe
  21. https://www.un.org/en/development/desa/population/publications/pdf/ageing/WorldPopulationAgeing2019-Highlights.pdf
  22. https://www.cbs.gov.il/he/mediarelease/DocLib/2017/138/01_17_138b.pdf
  23. https://data.oecd.org/pop/elderly-population.htm
  24. https://www.nature.com/articles/d41586-022-00074-x#ref-CR5
  25. https://www.nature.com/articles/d41586-022-00072-z?proof=t%29
  26. https://www.gov.il/BlobFolder/reports/licensed-inpatient-hospital-beds/he/files_publications_units_info_beds2022.pdf
  27. https://govextra.gov.il/media/16435/elderly-care-covid19.pdf
  28. https://www.mckinsey.com/industries/healthcare-systems-and-services/our-insights/hospital-care-in-2030 
  29. https://www2.deloitte.com/content/dam/Deloitte/global/Documents/Life-Sciences-Health-Care/us-lshc-hospital-of-the-future.pdf
  30. https://ieeexplore.ieee.org/abstract/document/8914198
  31. https://ieeexplore.ieee.org/document/7363546
  32. https://www.engadget.com/softbank-stopped-production-pepper-robots-032616568.html
  33. https://www.therobotreport.com/united-robotics-group-acquires-softbank-robotics-europe/
  34. https://www.unitedrobotics.group/products-services/hardware/
  35. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7840274/
  36. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8043496/
  37. https://fredericksburg.today/2021/12/12/mary-washington-hospital-welcomes-moxi-the-robot/
  38. https://www.bloomberg.com/press-releases/2022-05-24/meet-moxi-christianacare-introduces-innovative-collaborative-robot-to-give-nurses-more-time-to-deliver-care-and
  39. https://www.nursingworld.org/~494055/globalassets/innovation/robotics-and-the-impact-on-nursing-practice_print_12-2-2020-pdf-1.pdf
  40. https://spectrum.ieee.org/how-diligents-robots-are-making-a-difference-in-texas-hospitals
  41. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7312924/
  42. https://www.healthleadersmedia.com/innovation/pharmaceutical-robotics-require-careful-assessment?page=0%2C3
  43. https://www.igi-global.com/gateway/article/228954
  44. http://www.endorse-project.eu/endorse-outlook/
  45. https://www.therobotreport.com/tiago-robot-pal-robotics-ready-two-armed-tasks/
  46. https://promo-bot.ai/client-stories/promobot-in-hadassah-medical-assessor-at-the-oldest-clinic-in-israel/
  47. https://tadviser.com/index.php/Project:Arena_STEM_(Promobot)
  48. https://www.ynet.co.il/digital/technology/article/hyb9svwof