Developing a Cloud-based long term care management solution that helps care homes and hospitals to become more efficient at providing better care.
I joined my family business in 2018 that manufactures, exports, and sells consumable hygiene products, including diapers, pads, and absorbent bed sheets. The concept of developing an IoT product began when my mother and I met our hardware partner, who developed a wetness sensor that can be attached to a specially made diaper. After prototyping and receiving some early traction, I have decided to invest in the project to design and develop a Smart Diaper System.
In 2 years, I have incorporated Wonderkin Limited and built a team in creating our first IoT product, Nice Care Smart - Incontinence Monitoring System. This product enables carers in nursing homes or geriatric hospitals to be remotely notified of the patient's void frequencies, accidental falls, sleeping posture, inactivity, and motions in real-time, effectively easing the caregiving burden.
The IoT solution consists of a wearable healthcare device (i.e. diaper sensor), a gateway (Hub), and a virtual server hosted on the cloud. The system operates with the BLE (Bluetooth Low Energy) and WiFi.
When attached to the sensing lines on the smart diaper, the Sensor senses wetness in the diaper and returns a value indicating the wetness degree. It uses Bluetooth to send data to the Hub. The Hub, which can connect to multiple such sensors, gathers the data from all sensors within its range and transmits data to a web server hosted on the cloud. The Server then processes the data and displays on the web portal and apps.
Our Product aims to benefit the elderly and people with disabilities who suffer from moderate to severe incontinence. It aims to reduce the physical burden of caregivers and the associated costs due to the reduction in diaper change frequency.
As the company's sole founder, I managed fundraising, product design and development, and recruitment.
I led the team of one in-house UXUI designer and three developers to complete the development of the IoT solution. I work closely with business partners and suppliers as we iterate the design and development of the system, and run pilots and trials with hospital groups worldwide (Japan, Australia, UK, Hong Kong, and the US). In the end, I brought in a technical partner to continue to scale the project in the North American region.
As we licensed the diaper sensor development and patent from our partner, our team focused on designing and developing the digital client interphases (Web Portal & the Carer App), and the IoT infrastructure that supports data transmission. We also engaged actively in refining the features of the smart diaper and enhancing the overall customer experience of the carers, the administrator, and the elders.
We started by building the prototype, i.e., the diaper sensor, a diaper alarm (Hub), and the consumable diaper pads. I worked with my business partner to incorporate carbon strips during the diaper manufacturing process and to further customise the mechanical design of the Sensor so it can attach securely to the diaper strips when clipped and be removed when the used diaper is disposed. We also designed a diaper alarm that can sound and change colours (from green to red) when the wearer thoroughly wet the diaper. It has a Bluetooth chip embedded within and connected to each Sensor.
The prototyping phase took three months to complete. We then used the prototype to run a 3-month pilot with 40 patients and 20 carer staff in a nursing facility in Japan to test the concept of the diaper alerting system. We reduced diaper changes from eight to five times a day using the prototype. We cut the overall diaper consumption by 37%, equivalent to saving UDS 46 per user per month only on diaper costs.
We learnt that the beeping sounds from the Hub were considered an annoyance to the facility, and they would need a different alerting mechanism. Carers also wanted to monitor statuses and receive alerts on one interphase in the nursing station instead of needing to walk from the ward to ward to look at colour changes of the diaper wetness alarm.
With a proven case study and value to the consumers, I have decided to raise funds and create a team to develop the entire diaper alerting system, including the digital platform.
While I was raising funds and applying for grants for the startup established in Feb 2019, I outsourced the MVP development, including the database, the Server, and the frontend of the digital platforms. I also hired a product designer to work on the UXUI of the digital interphase. The entire development process took around five months, costing only ~ USD 38,000.
Designing a satisfactory customer experience for this IoT solution is challenging because the Product essentially serves multiple stakeholders, including the carers, the patients, and the system operators. Each user group has very different needs and sometimes has conflicting priorities. As we worked in a regulated market, we also needed to navigate the long sales process and get buy-in from government officials under the Hospital Authority and The Hong Kong Council of Social Service.
Due to all these design complexities, we continuously gathered early testers. We ran pilots with the prototype to understand how we can address pain points in the care homes and geriatric hospitals in Hong Kong. I have visited many care facilities, talked to thousands of carers and patients, and learned a lot about what care work entails at a facility. The user research helped me further refine the solution's design that can cater to the use care for a care home.
4. Carers usually have low educational backgrounds and usually speak non-native languages, which means the design of the interphase and call-to-actions must be visually intuitive.
5. The management & physician relies on paper records marked by carers to track diaper changes and incontinence trends (urine or stools) per individual patient. The process is manual and time-consuming. The system that centralises data like incontinence trends and diaper change timestamps would be a value-add.
6. Elder accidental fall is a huge care problem. When an elder accidentally falls, there is a chance that the seniors result in both physical or mental damage. Due to the emotional trauma experienced, the elder would become less mobile after the accident and further expedite the deterioration of muscles, increasing the risks of falls and other health issues. Falls without immediate attention usually result in fatal consequences. Fall detection or even prediction is life-changing for carers.
One of the challenges of working on this project is to develop a system consisting of both hardware and software. The communication protocol between the hardware, software, Server and databases has to be optimised. After completing several development iterations, we conducted a controlled test with 20 connected systems for two months to stress-test the durability and the system's overall performance.
We discovered that 3 out of 20 hubs automatically switch off after running 24/7 for around 45 days during testing. In addition, the Sensor's battery life only lasted for an average of 12.5 days in our controlled test, far from the 60 days that was expected.
We have communicated these issues to our hardware partner and supported debugging. We figured out that there was a bug in the algorithm implementation in calculating the voltages of the battery. We have also updated the Sensor's firmware by disabling some lower priority sensing computations to increase the battery life of the diaper sensor. As a result, we enhanced the performance of both the hardware with upgraded versions for further experimentation and trials.
After collecting more feedback from nursing homes and families, we have created multiple versions of the smart diaper and locked down the core features of the digital platforms.
The carer app notifies carers to take action on urgent tasks. It is designed for carers to navigate bigger nursing homes and complete care tasks on the go.
The portal provides an overview of all the elderly in the facility with detailed information regarding their diaper wetness level, accidental falls, posture, etc. Other features include
From Q4 of 2019 onwards, we continued running pilots externally. The go-to-market regarding the promised pilot projects was delayed due to the COVID-19 pandemic in Q1 2022. Some clients had to put the pilot on pause. Nevertheless, we still managed to run 15+ pilots with 200+ carer staff and patients in Hong Kong, New South Wales in Australia, London in the United Kingdom, Michigan State in the United States, and Oklahoma in Japan in a year.
We have measured a 25% reduction in diaper changes and wastage in one deployment.
By using the Carer app, carers and management has eradicated traditional paper recording of incontinence status. The system also helped reduce caregivers' physical burden by reducing the frequency of diaper changes, especially during night shifts.
To further scale the business and deploy the system in more nursing homes, the team needs to rewrite the Server's infrastructure and remodel the entire data transmission process.
During some of our deployments, we found out that the increase in the number of users using the system quickly lowered the performance of the Cloud Platform. In multiple instances, the diaper volume was inaccurate, and there was a significant data loss of incontinence volumes.
Data loss is an issue related to a massive load of data accumulated in the Server for processing every second. Thus, reducing the influx of data packets sent from the Hub to the Server and enlarging the server computation power is paramount for scale. Firmware update of the Hub is required to change how it processes data to reduce load to the Server, for example, reducing the data send frequency, filtering out data sent with low Bluetooth signal strength, and not sending historical data.
Other enhancements we identified that are of high priority:
By the end of 2021, I sold the Product to the US partner as it made sense to join forces with their technical team with IoT expertise to take it to the next level. Together, we created momentum with the pilot in the United States. The aspiration timeline was to launch in an Elder Home Group in 3 months where the system could support 120 people online simultaneously. In the following six months, the system could help 900 people.
About crossing the chasm - achieving product market fit
Removing a function in our system improves the Product for the user - the value proposition becomes more precise, and the workflow becomes simpler. At the early stage of the prototype development, we included temperature sensing as one of the features because the chip inside the Sensor has such functionality from the manufacturers' specification. We decided to enable it to utilise the chip we procured fully. Yet during user testing, we realised the temperatures indicated in the App created a lot of queries from the carers. They do not signify a person's temperature but rather the Sensor's temperature. We then realised that the temperature sensing function was unnecessary, and we decided to remove it from the App entirely.
Identify and focus on the quick wins
As a startup, the speed to market is paramount. Throughout designing and developing the Product, our team had switched focus several times and have delayed the process of achieving product-market fit.
We started developing a B2B product for the Hong Kong customers, then creating a B2C product, and later selling the Product to a different target segment, i.e., the parents and babies in China. Between March to June 2022, we even started selling personal protection equipment to our business partners to cope with the COVID-19 outbreak. Reacting quickly to changes in the commercial environment is not wrong (In our case, it helps generate cash for a longer runway), but we deprioritised the efforts in developing our core product where the sustainable business potentials lie.
Understanding the market is critical to prioritising our development efforts. Each country or jurisdiction has their regulations concerning healthcare systems. For example, in the US, carers cannot walk around the phone as it is against HIPAA. Different certifications and licenses are required, e.g., the UK requires MHRA, and the US needs FDA approval.
As the requirements per region differ, it is a lot of effort to customise the Product to all different segments and markets. There are generally two ways to work around this: (i) Designing a system that could easily be customised or "white-labelled" by clients; (ii) Designing a customised system for a small market and focusing on winning in this market.
If I had to start all over again, I would first identify a segment I want to tap into that would reap the most business potential. I would focus on establishing a case study that could help generate more traction and referrals, then use this example to raise funds to scale production and business development later on.
Steven Jobs once said
"People don't know what they want until you show it to them. That's why I never rely on market research. Our task is to read things that are not yet on the page."
In this startup experience, I fully understood what it means by that. An actual need can only be validated in a real-life scenario. While we were trying to collate feedback and requirements early in the process, care homes and facility managers said they wanted to see centralised data to develop insights to help streamline the care home operations. They said they would value having a Web Portal more than using a mobile application or relying on the Diaper wetness alarm to notify of a diaper change.
However, as we rolled out the system to the care homes, very few looked at the Web Portal. Most of them downloaded the App on a tablet and used that as a single source of truth for all alerts and historical records. Based on the feedback we obtained after the trials, we realised most carers only relied on the Diaper Wetness Alarm (Hubs) because the Tablet is a shared tool among all carers.
In many pilots we have run, we put 100% effort to ensure the success of the system adoption and delivery. In one of our corporate partnerships in Japan, we invested over six months having weekly meetings, monitoring and reporting the pilot results, and even localising the system to adopt the Japanese language.
Unfortunately, we discuss how to formalise the commercial agreement among the parties involved only after the pilot's completion. There were contractual terms that we could not agree on, and in the end not able to do business and reap any return from the investment.
Designing an IoT solution is not just about creating a system but designing an experience and the business process that helps the users use the Product.
In one of our deployments at the geriatric hospital in Hong Kong, I visited the nursing homes ten times before getting the sales order. I sat outside the hospital on the floor for hours, trying to connect to the hospital WiFi and configure the system as we could not enter the nursing station due to COVID-19 restrictions.
During the sales process, we looked at floor plans of the nursing homes to suggest the number of hubs required to cover the entire space. We create multiple subscription models for different care homes and hospitals. We factored into the costs of the network and even had to partner with the network service provider to offer 1-year or multi-year plans. We even went as far as coming up with proposals for nursing homes to apply for government grants to upgrade their systems to make it easier for them to adopt our system. To close the deal, I had to include extra batteries for the sensors, a network plan and work out the entire workflow for the hospital to take it forward.
Developing a commercially viable product also means that I have to develop training materials and reference guides for care home management. During my tenure, I had to deliver training sessions for the carers, provide extenders and routers, take a 15 hour trip to one nursing home in Australia and stay onsite for a week debugging issues to ensure the system worked.
I was the 'CPO', 'CFO', 'CEO' and 'CTO' of the company. I raised funds, liaised with partners, consolidated requirements, synthesised user stories, and managed the development with in-house engineers and outsourced software and hardware companies. I had previous experiences with app development and cloud technology from my past working experiences. Still, I have underestimated what it takes to develop a fantastic IoT product or become an IoT company.
Due to the lack of knowledge of IoT, I encountered many issues and uncertainties with hardware management and communication network/infrastructure. There are too many unknowns that I did not know of initially, resulting in misplanning of budgets and timelines. I've always known I need a technical partner but finding one is easier said than done.
Visit www.wonderkin.co to learn more about our products. If you happen to be interested in collaborating, please get in touch with me about my Product.
Energised by bold ideas and lateral thinking, I’m up for solving big problems and transforming the world. Count me in for innovating products, processes, and customer experiences.
