Introduction by Zach White, Insights Manager, GSMA

In this guest blog from Mobile Alert Toilets (MAT), Joyce Klu (MAT founder and CEO) and Denzel Wamburu (MAT data engineer) take a close look at how sensor technologies are advancing and how these can be used in sanitation services and businesses. The need for, and potential of, sensors to improve sanitation service delivery is clear. By monitoring a set of ‘distributed assets’ (the toilets) cheaply, safe sanitation providers can optimise collections and, in doing so, lower costs. Currently, the costs of safe sanitation provision and customer willingness and ability to pay are not well aligned. Additionally, not disposing of waste safely (dumping it in drains or water bodies) can be cheaper for providers. This gives rise to the ‘pick two’ problem in urban sanitation. In many cases, one of three areas – profit, safe disposal, and serving low-income customers – usually has to be sacrificed.

Source: Developed by GSMA staff in discussion with Ian Ross

While not a cure-all, technologies that contribute to substantially lowering the costs of providing safe sanitation services will go a long way to resolving this problem. At the GSMA, we last looked closely at the potential of sensors back in 2015 through a grant to Sanergy and SweetSense. At the time we found that sensors to optimise collection did indeed have operational benefits, but that the technology needed further development, and the costs were too high for their use to make sense, given cheaper solutions. Nonetheless, others have continued important experimentation in sensor placement and technology development to reduce the costs. Combined with improved connectivity, these developments mean that GSM-monitored toilets in low-income areas are close to the point where they may make good business sense in many contexts. Additionally, sensors have a vital role in monitoring public health risks. In many urban areas, it is common that faecal waste is discharged straight into drains, water bodies, or just the streets. Monitoring when toilets are close to being full and ensuring waste is collected supports protecting public health.  

We’d like to thank Joyce and Denzel for preparing this blog. It grew out of a conversation we had at the end of 2020 about their work. We were interested in how they approached developing sensors and integrating them in their operations, we hope you find it as insightful as we did.

Mobile Alert Toilets (MAT)

MAT is a social business, that improves overflowing and unsafe pit latrines in schools and communities in the urban slums of Kenya by the use of sensor technology.

In Kenya, approximately 6.4 million people live in urban slums with little or no sanitation infrastructure. Korogocho slum, with an estimated population of 200,000 people, including school children, either do not have sanitation facilities or rely on overflowing pit latrines. The ever-growing population in the slums and lack of planning for sanitation infrastructure has made the situation dire for both schools and community.

Dealing with overflowing pit latrines, smells, flies and worms greatly impacts the performance and attendance of school children as well negative health implications for the entire community.  Additionally, the method for collection and disposal of faecal sludge from the latrines involves manual scooping and dumping into nearby water bodies- An estimated 140m³ of faecal sludge is dumped into the nearby Nairobi river everyday rendering it unsafe to use. The practice of scooping is not only unhygienic to sanitation workers but also threatens public health and the environment.

In response to these sanitation challenges, MAT renovates/reconstructs existing dilapidated latrines by re-enforcing slabs and improving the super structure (roofs, walls, ventilation). Additionally, the latrines are fitted with containers (only in areas where exhauster trucks cannot access) and ultra-sonic sensor to monitor fill levels and detect overflows and send automated messages alerting on timely emptying and safe disposal.

MAT’s use of sensors in our operations

In each of the latrines constructed, an ultra-sonic sensor is placed below the slab to read the fill levels of the pit. The sensor node is made up of a waterproof ultrasonic sensor, a microcontroller, a battery and a SIM module. The ultrasonic sensor measures the fill level of the toilet, pinging depending on the schedule. The wake-up schedule is auto-configured based on the fill rate of the toilet. The higher the fill rate the more frequently the device wakes up to send data.

Once the distance is acquired, the data is tagged with the toilet ID and is sent to our cloud software over the SIM module. Once the operation is done, the sensor node goes to low power and sleeps until the next operation cycle. When the reading indicates about 90 per cent fill level, exhausting/collection services are scheduled and waste is collected and then disposed safely.

The current challenge with the sensor is its battery life as it needs to be recharged/changed every six months. However, we are consistently researching how to improve the efficiency of the sensors and hope to be able to achieve a two-year battery in the future.

The sensor market

Off-the-shelf sensors designed for underground conditions in pit latrines are not readily available in the market. The ones already in the market are costly, bulky and still require modification and software integration by an expert. Market costs for a sensor that collects distance data and transmits it is about $1,000. Buying off the shelf and customising is harder than building one from scratch.

Our approach to install sensors below the ground poses several challenges. The sensor should be water proof, should withstand acidic conditions, connectivity should be available below ground level and should be tamper proof. For these reasons, we have had to build our own custom-made sensors and software. Our hardware and software cost per unit is relatively high due to fact that we are currently operating at small scale. At the moment the sensors we build cost roughly $145 per unit. However, with scaled-up production, we think the cost could reduce to as low as $30 – 40 per unit.

Before switching to GSM technology, we were using low-power wide-area network technology provided by SigFox. This proved a challenge since the network did not cover all areas within the city and reception was poor indoors. In addition, it required an annual subscription fee which was quite costly. Our plan is to move to LTE-MTC low power wide area (LPWA) technology to facilitate longer battery lives.

Over the years, battery technology has improved drastically. Batteries are cheaper and more compact. We are working to include a charging mechanism to avoid frequent replacement of batteries. Due to the small scale of sensors we are producing, the cost is high but once we bulk produce, the cost and size of producing the units will reduce.

We use an ultrasonic sensor to measure the fill level. This works by sending out a sound wave at a frequency above the range of human hearing. It is low cost but inaccurate. To mitigate this, measurements are taken five times to filter out the inaccurate measurements. We are researching how to improve this using laser sensors.

The opportunities in safe sanitation provision  

Before trying to introduce sensors into sanitation operations, there are key things to consider. Firstly, the availability of the sensors in the market and/or a technical expert to build one that can suit specific latrine conditions. Secondly, conducting a cost-benefits analysis is crucial as sensors can be quite expensive. In MAT’s case, the benefits outweigh the cost, not in terms of profits but in terms of environmental protection and reliability of service provision. Thirdly, it is important to ensure the security of the sensor depending on the location where it is installed. This can be done having it completely underground with no visible cables to protect it from vandalism or theft. Finally, the availability of the network for data transmission has to be guaranteed at any given time.

Since beginning its operations, MAT has constructed a number of latrines both with and without containers. Our containers have capacity of between 250 – 300 litres while the size of the pits are unknown since we are not the one that built the first latrines. Sensors help with detecting an overflow in both scenarios. Though one may argue that if you know the size of the container and the number of users, you can automatically calculate the fill rate of containers and pits, experience has proved that this is not usually the case, particularly in the slum setting where populations are unpredictable. Fill rates vary from time to time and simply relying on calculations may result in overflows and/or an underutilised collection schedule.

Therefore, safe sanitation providers in such settings should endeavour to incorporate the use of technology such as sensors for optimum operations. Additionally, in the slum setting, there is a conglomeration of houses without any street address and locating latrine facilities can be a tedious task. The sensor technology not only detects overflows but can also incorporate GPS for mapping locations of the facilities. This enables efficiency in collection through planning of routes.

As the world is moving rapidly towards technology in all aspects including the provision of basic services such as water and sanitation, it is crucial for all stakeholders in the sector to embrace change. Local authorities, in particular, should prioritise partnerships with entrepreneurs and tech companies involved in service provision of this nature as well as provide enabling conditions for the success of its implementation. There should also be incentives such as tax relief on hardware to encourage pilot testing of novel technologies such as the use of sensors in latrines. This will not only encourage start-ups but also help authorities in mapping and understanding the proportion of the population with basic services. Funding entities/donors on the other hand should consider making smart sanitation technologies one of their priority areas of funding not only because sanitation businesses cannot survive without blended financing especially when serving the vulnerable but also, new technologies require a lot of research and trial and error which can be costly. 

---

About the authors

---

The GSMA Mobile for Development (M4D) Utilities programme is funded by the UK Foreign, Commonwealth & Development Office (FCDO), and supported by the GSMA and its members.