Five questions with SmartClean

Keeping public toilets spotless can be a huge challenge, especially given the shortage of cleaners and the high turnover rate in the cleaning industry. Enter SmartClean, a Singapore start-up developing technology which makes use of sensors to monitor toilet cleanliness and can send alerts to cleaners in real time. The company graduated from AIRmaker - an accelerator that grooms IoT (Internet of Things) start-ups - last month and is in the process of testing its solution.

How did SmartClean get started?

Ms Aw I have been running a cleaning company called Spotless, which I started about six years ago with my brother. We mainly do commercial cleaning contracts at offices, factories and shopping malls. We have about 35 employees and 26 cleaning contracts.

The cleaning industry faces a high turnover rate and manpower-related issues, and I have been looking for ways to make the company more productive and allocate manpower more efficiently.

Mr Lav Agarwal The other three of us are PhD (students) in electrical engineering at the National University of Singapore and Nanyang Technological University. We did our undergraduate studies in India and came here for further studies.

We three already have a two-year-old company called Synapse Technologies which offers robotics solutions for different use cases.

We wanted to apply this to the cleaning industry and try to solve some of the issues it faces.

 

Q How does your solution help overcome cleaning industry challenges?

Mr Lav Agarwal In Singapore, for example, most of the cleaning manpower comes from nearby countries and or are older people. So if you can optimise cleaning operations, the workload can be reduced.

Our solution involves installing a set of sensors in toilets which evaluate different aspects of cleanliness in real time. This includes air quality monitoring, which means checking if the toilets are smelly or not, and wetness monitoring, which checks if the floors are wet or dry.

The sensors can detect anomalies and alert cleaners in real time, so it becomes like an automated cleaning supervisor.

This will help move the industry from schedule-based cleaning to need-based cleaning.

We also put in people counters to capture user patterns as well as touchscreens which can capture user feedback about the facility.

Cleaners can also automatically log their tasks in the system.

 

Q What was the development process like and what is the next step?

Mr Lav Agarwal The first version took about 10 months to develop. The next step for R&D would be to develop artificial intelligence, as well as further improve on the sensors.

The cleaning industry does not require much training or a skilled workforce – people come and go very often.

This is why we’re also building an AI (artificial intelligence) which will allow the “cleaning supervisor” to give specific instructions instead of just detecting anomalies.

The instructions given can be clear enough so that the system is not dependent on a specific cleaner.

In future, instead of deploying cleaners to the facility, a cleaning team can be located in a smart city serving multiple locations at once by receiving alerts.

We are also working to improve the aesthetics of the sensor modules, as they will be deployed in public places like toilets and airports so they need to look good.

Infrastructure owners will always have concerns about the reaction of the end-users when they see the sensors installed – whether there will be wires running around the premises and so on.

All the sensor casings were 3D-printed and the designs will be iterated and improved upon for production in future.

 

Q Have you tested the solution?

Mr Lav Agarwal We’re deploying it at a few facilities in a pilot. One is in the men’s washroom at Bash (Build Amazing Start-ups Here) – a start-up facility at Block 79, JTC LaunchPad @ one-north.

There are also plans to test-bed the solution at an Ascendas facility – AIRmaker is a joint venture between Ascendas-Singbridge, SGInnovate and China’s Runyang Group.

But we still need more data to train and develop the AI. At the moment, what is being tested is data capturing and sending alerts based on information received from the sensors.

 

Q What are your growth plans?

Mr Lav Agarwal We are in the process of raising $350,000.

The money will be used for R&D and product development, as well as hiring more people for the team.

Photo:KEVIN LIM

Share This Post

Share on facebook
Share on linkedin
Share on twitter
Share on print
Share on email