A.B.U: Accident Broadcast Unit

An accident broadcast unit and smart caregiving system aimed at improving the quality of life of the elderly population.

There is no safer place than home. It seems like an undeniable truth for many, but that fades away as we grow older.

10% of Argentina's overall population is over the age of 65, that’s over 4 million people according to the country's National Institute of Statistics and Censuses, an age group prone to domestic accidents. Furthermore, one out of five of them lives alone.

The World Health Organization notes that falls are the second leading cause of accidental injury deaths worldwide and identified that the elderly are at the greatest risk of serious injury from falls. In Argentina, one out of three senior citizens falls during a year, with half of them doing so more than once. According to the National Survey on Quality of Life of Older Adults around 50% of falls in the elderly result in bone fractures.

Falls are a public health concern we should address, especially among the aging population.

Another very important source of injury is fire and gas-related incidents. Who hasn’t heard of a stove left on? Or a gas heater consuming all the oxygen? In Buenos Aires, Argentina’s capital, over 80% of intoxications are related to malfunctioning or misused gas appliances. This problem becomes more severe with age, due to the natural memory loss in elderly people.

Something else that can easily be observed whilst engaging with an older person is the difficulty they have using a modern device, like a smartphone or a tablet. There is no easy way for them to get a grasp of all the technologies available to the younger generations. This usually leads to frustration and alienation, and sometimes even some form of recrimination by their children, related to not picking up the phone, or misconfiguring it. 

The concern we share about all these issues and the love we have for our grandparents (or as we call them in Spanish - abuelos) brought our imagination to action, and even better, brought us here, to propose some solutions to these problems and give them their dignity and independence back.

A.B.U. is an innovative system, merging several well-known technologies, alongside cutting-edge innovations, to provide a simple, intuitive, and yet very effective solution to all the above-mentioned problems. This system is designed in a way that the elderly population feels comfortable using, and that enhances their life quality. It consists of several parts, wirelessly connected:

• Base unit: mains powered with battery backup wall-mounted or desk-placed device, with a screen with big readable fonts and a simple UI, a strong speaker, and omnidirectional microphones. The device should provide simple voice-activated services, like current “smart home” devices, but with extra features tailored to the elderly generations, like advanced emergency calls and fall alerts to an emergency contact or first responders, and some daily activities helpers, like weather warnings and medication reminders.
• Fall detection cameras: simple and cheap cameras, placed in rooms of interest, in order to detect falls. These cameras have no connection to the Internet, just to the base unit, and their only task is to perform the optical detection of a fall. No recording is necessary, all processing is done on the fly.
• Fall detection RF sensors: inexpensive sensors that talk to the base unit, used to detect a fall, based on RF signals, with no optical detection, providing enhanced privacy. 
• Smart wearable device (optional): should the elderly wish to do so, a smart wearable device may be connected to the system to monitor vital signs and log useful information.
• Smoke, gas, CO, and CO2 sensors (optional): small sensors used to detect hazardous situations, placed in critical spots around the house.
• Companion App: meant to be used by the caregiver, allows the configuration of the system remotely, and vital signs monitoring if a wearable device is being worn.

Our solution aims to improve the life quality of elderly people who live alone or are in need of constant supervision. One of the many ways that we as a society mistreat the elderly is by neglecting their independence. In a lot of cases, this is because they fail at some of the tasks in day-to-day life, may it be taking the right medication at the right time or being able to read a text on a new smartphone, or they make us worried about their safety. This is especially true with older adults, who often have a poorer sense of balance, and/or some mobility issues. At that age, bones are more fragile than ever, and a simple slip can result in broken bones with life-threatening consequences. 

Let’s say the elderly under care falls down. This is where A.B.U. kicks in. The event is detected by either a fall detection camera or an RF fall detection sensor and relayed to the base unit. Depending on the user-defined configuration, an alert can be sent to an emergency contact. This can be done via text message, a phone call, or even a video call. The system could also be configured to send vital signs information, if a wearable device is being worn, to the emergency contact, or even to a predefined hospital or medical center. This would allow for a quick response should urgent care be needed. Similar alerts may be generated if another sensor detects a hazardous situation, like an excess of carbon dioxide or a gas leak.

Being able to give them their independence back is a highly relevant matter, because they are still human beings, loving and caring. Having this technology would enable them to be confident, to solve their own problems, and to feel they are cared for, but not constantly watched and judged. This technology enables them to not feel like a burden, but to feel like humans again. 

Our team has two very distinct profiles. On the one hand, Martina is an advanced biochem student. She is a volunteer for the Red Cross and has lots of knowledge about medical-related topics, crisis management, and vulnerable communities. She has a passion for new and innovative technologies regarding science and health. Juan, on the other hand, is an advanced electronics engineering student. He has brought to market a few IoT products, designing both hardware and software, and enjoys the challenges associated with the manufacturing process.

This combination brings the team a unique perspective, having knowledge of both sides of the problem: human and technical. It also provides us with the right connections to make this project happen, having contact with the manufacturing industry and with healthcare centers.

We, as grandchildren, are inevitably affected by the reality of our grandparents, and eventually, the future of our own parents. 

During one campaign as a volunteer at the Red Cross, Martina had the opportunity to connect with older adults at retirement homes and gain insight into their most pressing concerns.

After the initial concept was drafted, statistical research was carried out to ensure the relevance of the problem and its scale in the actual population. The results were compelling enough to spur us on to further develop the idea.

Once the system is launched, ongoing investigation and fine-tuning of its features will be necessary to ensure its effectiveness in addressing the problem.

One of the key points of our solution is the integration of several technologies. Voice-activated assistants are common nowadays, and so are IoT home sensors and surveillance systems. However, these are all separate systems, with no integration. Of course, one can do home automation with tools like Home Assistant, Node Red, and Tasmota, but that’s not something the general public is trained to do, let alone an elderly person. Our solution is to integrate systems, to make them accessible to someone who doesn't know how to use them. It needs to be intuitive and inclusive. 

Another significant portion of our solution is the RF fall detection system. This enables the user to have complete privacy and peace of mind, knowing he is not being watched. This technology is a game changer, and will surely impact several markets in the following years. Being one of the early adopters is crucial to move the technology forward, and to help as many people as possible in doing so. The uses are practically unlimited. From estimating dangerous actions in workplaces to making sure gym exercises are done with the correct form, the possibilities are endless. 

One more important feature of our system is the absence of an obligatory wearable device. Many fall detection systems are based on accelerometers and gyroscopes placed on a wristband. This approach works very well but relies on the user actually wearing the device. In our system, this is an optional feature. However, such devices are really useful to monitor overall health, like heart rate or blood pressure, thus our proposed system would allow them as optional accessories. 

As previously mentioned, no recordings are necessary to perform optical fall detection. Nevertheless, the system may be configured to start a video call right after a fall is detected, and send a short clip of a few seconds right before the incident happened, to allow the implementation of future preventive measures.

1) Decrease the necessary amount of time for caregivers to respond to accidents.

2) Minimize the potential for additional injuries that may occur after a fall while trying to reach for help.

3) Increase the number of preventive measures implemented after video clip analysis.  

4) Improve the reliability of fall detection in senior citizen homes.

5) Improve adherence to treatments and routines thanks to reminders.

6) Achieve a record of measurements and vital signs for statistics.

This project heavily relies on advanced technology.

One of the core ideas is the RF fall detection. This is based on WiFi pose estimation, a novel approach to infer body positions based on phase and amplitude changes of known signals in the WiFi frequencies. This idea is discussed in depth in “DensePose From WiFi” by Jiaqi Geng, Dong Huang, and Fernando De la Torre (https://arxiv.org/abs/2301.00250). Traditionally, this is done using optical recognition, which means using cameras. For the WiFi pose estimation, several WiFi transceivers should be placed around the rooms of interest. For optical recognition, cameras must be placed instead. Either way, the result is the real-time position of the person in the room. 

Having this real-time position of the body makes it easy to detect a fall. For example, if the person was standing up, and within a short time period (for example less than 250ms) the body is kneeling or lying down, a fall has occurred and the alert protocol is activated. Of course, a more complex algorithm may be used to detect a fall, even an AI trained on lots of fall scenarios, and normal conditions, to improve accuracy. 

Another core technology is the voice control over the system. It has to be simple and effective. This is a well-researched topic, and there are many implementations to choose from. It enables ease of use and intuitive commands. For the elderly who have speech or hearing impairments, a touchscreen with big readable fonts and an extremely simplified UI is proposed. 

CO, CO2, gas, and smoke detectors are common, and integrating them with our proposed system is not complicated at all, but provides enormous added value.

Currently, our solution has not yet been launched and we do not have any numbers to report in terms of the number of people we are currently serving. 

Next year, we plan to design and develop a prototype of our solution. Then we intend to conduct pilot tests with a select group, we could be talking about the first 50-100 homes for trial. This would allow us to gather data and feedback to fine-tune our approach. Based on this information, we will be able to adjust and optimize the integration between systems and the hardware design. 

Once everything is set up, we will then be able to provide the service to a larger community. As per the statistics previously mentioned, there are more than 4 million people over the age of 65 in the country, of whom nearly a million live alone and could potentially benefit from our product one day. Considering the uneven access to healthcare in our country, that number is likely to be lower. Therefore, our goal is to work towards reducing accessibility barriers and establish partnerships and agreements with organizations to enhance our reach and impact.

An increasing elderly population and the aforementioned issues are not only restricted to Argentina, but transcend frontiers. Conducting market research to understand the needs and regulations in other countries is key. Our solution could be expanded to other countries in Latin America and the rest of the world to reach a larger population. Not only would it increase the impact we make, but also will open new markets and generate additional revenue. This will present the need to partner up with local organizations and companies to help navigate the different healthcare systems and regulations, as well as to provide localized customer support such as multi-language options.

The main barriers for next year are mostly technological ones. As we need to make a functional prototype, some of the main considerations to keep an eye on are: system integration, WiFi pose estimation, and hardware design.

System integration, often referred to as “integration hell”, is the process of making several, often very different, systems work together seamlessly. In our case, this includes making sure the voice control works alongside the text messaging backend, the alarm and alert system, the fall detection mechanisms, the phone app used to configure the base unit, and so on. This is usually not trivial, but it’s not impossible either. It just takes time and effort.

The WiFi pose estimation is a big challenge, given that the paper this technology is based on was published rather recently. Nevertheless, we have access to the necessary test equipment and advisors to make this come true. Should this take longer than expected, the system can be used only with optical detection, which is a tried and tested method.

The hardware design is somewhat complex, but not impossible to do. We feel confident in our ability, given our past experiences, to deliver a good prototyping unit. Something to keep in mind is the component selection, not because of their cost, but rather their availability in the market, given the recent chip shortage induced by the pandemic. 

Currently, we are not working with any organization given the fact that we are still in the stage of developing the concept.

As we mentioned, working toward our goals means establishing partnerships and agreements with organisms at some point to expand our reach and impact. This may include working with healthcare providers, government entities, and community organizations.

Not oblivious to the unequal health access, and despite our additional efforts to decrease the costs required, collaborating with them could include offering financing options, subsidies, donations or even working with insurance providers to cover the costs of our device. 

We intend to provide our solution regardless of location or socioeconomic status.

We plan to adapt the solution to a for-profit social enterprise model, achieving economic value to support social impact. This means the organization is an enterprise embedded in the social program, being the purpose of the business to solely serve the clients as its main reason. 

Our main strategy is to sell the base unit as a product, with a one-time payment, with assured updates and maintenance. Additional sensors and cameras will also be sold as a product, taking into account that every house will need a different quantity, depending on the number of rooms to cover. There might be future options for additional paid services, but the main objective is to guarantee full performance with the first investment to ensure long-term affordability. 

We are considering targeting three distinct client types: 

1. Financially self-sustainable end user: sell the device directly to end users, and offer additional paid services and customizations. 

2. Financially unsustainable end user: provide access to those who wouldn’t otherwise afford it, via third-party payers such as healthcare providers, government entities, and insurance providers.

3. Private institutions: provide a service as a multi-user system, adaptable to facilities such as retirement homes and geriatric hospitals, with heavy customizations and large deployments.

Even though at first we may rely on fundraising through grants, donations, and crowdfunding to help cover the initial costs of development, in the long run, we plan to generate enough revenue to cover our expenses and reinvest to improve and scale our solution. 

Our team has some personal resources ready to invest in the idea and start prototyping. This initial phase will be also alleviated by our connections, having good relationships with local and international manufacturing firms, willing to work with us as we are regular clients. In this prototyping phase, we hope to end up with some functional systems, ready to show potential investors the true value of our solution. 

Should it proceed according to plan, the next phase is to manufacture 50-100 units to test in volunteer homes. A viable option is to offer them the units for purchase after the trial at production cost if they liked how they performed, or return them with no charge. This would allow us to get some cash flow for the next phase, and also some honest feedback, given that if the system is actually helpful, and people are willing to pay for it, we are on the right track. 

After this small test batch, and if the product is successful, we would re-evaluate our business strategy to ensure its robustness and scalability. This will include assessing the costs, revenue potential, and manufacturing process.