EyeKnow

In Israel and around the world, we find ourselves too often shocked again by the news of yet another case of child abuse, such as the Carmel Mauda story that has shocked has the nation, or the death of baby Yasmin Vinta by her caregiver, along with more horrific stories of repeated elderly abuse, such as in the case of violence in the Neot Kipat HaZahav nursing home in Haifa, a few years back.

Social, demographic and technological changes that have taken place in recent decades, worldwide and in Israel particularly, have led to a steady increase in the amount of helpless individuals, on both ends of the age spectrum: a natural increase in the elderly population, and on the other hand a "Baby Boom" - a massive increase in the number of infants aged 0-3, who spend most of their early days in nurseries and daycares, mostly privately owned, where their safety and wellbeing are left in the hands of strangers, with no actual means of monitoring what happens, or any kind of parental control.

Research has shown that these age groups are the most likely to suffer from improper behavior, violence and neglect by their caregivers, due to the inability of infants or the elderly to protect themselves and/or report what is done to them, which makes them more vulnerable than any other group, to risks to their wellbeing and safety.

According to estimates, 1 out of 5 elderly people in Israel experience some form of violence by caregivers. Worldwide, according to statistics by the World Health Organization, 2 out of 3 caregivers have admitted to having used violence towards the elderly in treatment facilities. The organization has also reported that 15%-30% of seniors in nursing homes or receiving home-care experience verbal or physical violence by caregivers. Causes to this harsh phenomenon include a nearly complete lack of means of supervision, leaving children and seniors, as well as their families, vulnerable and helpless.

Having to physically arrive at care facilities, to examine their compliance with regulation, is a major problem which prevents authorities from being able to continuously monitor this compliance, and allows for large "windows" of possible violent events, which go unseen. A survey conducted by EyeKnow, among children of elderly parents in need of care, has found that violence by the caregiver responsible for the parents' wellbeing, is one of the two biggest concerns that worry them the most.

A survey conducted by the company in collaboration with survey expert Geocartographia, has shown that around 40% of those surveyed have said that the EyeKnow service is relevant to them, a number which rises to 47% among those who have a close elderly family member or are over 70 themselves, while 71% find the technology to be innovative and unique, compared to current solutions:

The Solution

Now, for the first time, EyeKnow is harnessing the most advanced technology in order to answer this profound and painful social need. The company is developing AI based algorithms and methods, for identifying, classifying and monitoring interactions between caregivers and the elderly or children under their care, with a main focus on multidisciplinary algorithms, pre-designed to suit diverse monitoring and control systems, for use among varied populations.

The company and its tech will allow for a new generation of detection, classification and monitoring solutions, based on AI, which have not been possible as of yet, aimed to enhance and boost the monitoring and training processes of professional staff members, in order to protect vulnerable individuals in various care facilities. This, while improving the care experience and quality of life of helpless individuals, and improving the professional efficiency of caregivers and facilities.

The technological solution developed by EyeKnow addresses a profound problem as shown by research - elderly people who experience violence die faster, and continued exposure to harsh environmental conditions and poor quality care is a critical developmental risk factor for children. The technology is designed to help create beneficial environmental conditions, at home or other care facilities, and help protect and enhance all factors related to an ideal emotional, social, and cognitive development - just like every adult, child, and baby deserves.

Additionally, due to the great public pressure in response to abuse reports in recent years, authorities and government entities have finally come to the understanding that a country must provide its helpless citizens, starting with the elderly and infant populations, with high-quality protection and supervision means and environments of care, which are suited to their needs, under continuous monitoring and evaluation, and according to clear and uniform standards.

Technology

So far, the company has developed a POC based on a unique algorithm for classifying a limited number of interaction types, between daycare staff and children under their care. The results of this initial algorithm, developed as a prototype, are in an ongoing process of upgrade and adaptation, while the company is also developing and adapting the technology for the elderly care market, to allow identifying, classifying and monitoring interactions between caregivers and patients who are at risk.

The company's tech is based on AI models which are trained using original and unique videos of violent actions, collected by the company in collaboration with its partners. These models are accessible to customers via cloud and web application. In addition, the company employs a control team, who verifies that system alerts are accurate (using a method called Human-In-The-Loop). This way, the company can provide very high accuracy results, with a rate of nearly zero "False Positives".

In the initial phase, the tech will focus on the following aspects:

• Building deep learning algorithms to identify and classify caregivers' violent interactions with the elderly and infants, based on visual data
• Establishing a "pipeline" to run algorithms in an operational environment, based on CCTV cameras, DVRs and IP cameras installed in homes and facilities, and later on an automatic analysis and profiling of behavior
• Creating a business intelligence (BI) system based on the collected and analyzed data 
• An automated system for improving the skills of staff members and streamlining the quality of care and operation of the care facility

With this campaign the company intend, among other things, to expand the technology and its capabilities:

• Upgrading the AI model to identify violence by caregivers towards the elderly
• Creating algorithms based on deep learning, for the detection of violence against the elderly for a production system at scale
• Running the systems and algorithms previously mentioned on cloud-based servers, as well as on on-premises dedicated hardware (prem-on)
• Development, implementation and support of EyeKnow's system with a launch planned in cooperation with one of the largest communications companies in Israel
• Building a GUI-based system based on the various analytics

How does it work?

The company's technology is based on and divided into software and hardware architectures. EyeKnow's software architecture is divided into 3 main parts:

1 . Interfacing with security cams and data collection
2. Processing data and running the algorithms
3. Sending alerts in case of aggressive and violent incidents to clients via an app

In fact, EyeKnow's technology interfaces directly with standard security cameras, and manages the reception of frames with the help of a controlled queue management system. The system has the ability to collect information from many cameras in different locations in the care facility/facilities, while fully adjusting and creating a clear hierarchy of the data stored, without any data loss.

The multi-layered system can connect to any camera. In the next stage, the platform uses deep learning algorithms and models to identify different behaviors, in order of importance and according to the development plan, including identifying aggressive actions, clear physical violence, beatings, kicks, pushes and pulls, etc.

The technology is based on a cloud-based inference system such as Amazon AWS, and also includes internal validation and labeling systems, which allow the team to verify that the events the algorithms identify are in fact real and require immediate attention, before transmitting the processed data to customers. The platform also includes a business intelligence system and a dashboard for convenient monitoring of all information collected, as well as analysis based on the data.

The system will be developed as an app and also as a web application, initially for use by children of elderly people in need of care, or parents of infants for home use, and later on also for use by owners or managers of daycares / nursing homes etc, with a built-in ability to view videos of the events reported by the platform.

Patents

EyeKnow intendsto register two exclusive patents, which represent an attractive and important IP, including an element that covers the ability to detect aggressive movements on which the technology is based, along with an additional patent which covers the profiling of caregivers who are likely to perform violent acts.