“Spectrino - Arduino devices that can be implemented on a wide spectrum of touch-free tinyML based housing and society systems.
The pandemic has introduced a constraint to social interaction: distance. Considering this factor of risk, countries all over the world have been in varying levels of quarantine, and many malls have had to close down due to significantly lowered consumer count. This has led to a very, very high level of layoffs of mall personnel, as well as similar economic challenges for business owners.
This has caused relative low-income (less than $40,000 in annual earnings) job loss levels as of July 2, 2020 in the US due to COVID-19. Accommodation and Food Services, as well as Retail Trade and Entertainment collectively count for ~4,000,000 of the estimated jobs lost.
The economic challenge faced by different industries, where Food, Consumer, and Retail come up in the Top 6 industries with the highest number of employees laid off (amounting to estimated 20,000+ jobs). Assuming “International” to include for all non-US areas taken into account by this research, the total global layoff count reaches an estimated 103,000+ jobs.
Considering the given data, the team has determined a major challenge to be that there is risk uncertainty, with regard to population density in different shops, for the malls that are still open. Aside from this, while wearing face masks and avoiding touch is mandatory in many places, there are still violations. This makes it more difficult for those mall personnel and business owners, who cannot afford to work remotely, to safely navigate this new normal workspace. This begs the question: how can we guarantee to a decent certainty level that, at any point in time, a particular shop in the mall is safe to enter?
We all are now fighting against the prevailing COVID-19 pandemic. And also, now we are in a situation where we have to adapt to the prevailing conditions with more safety measures. While life coming back to normal with more safety measures to avoid virus infection, adding safety within the public places and crowded areas are also prevailing in the cities. But there were many situations where we have to break the safety measures and interact with an unsafe element to meet the needy. Here, the project is dealing with the prevention of COVID-19 spread though touch interactions or touches.
Hence I decided to automate the most commonly used devices in housing and societies to ensure hands-free communication with devices.
The following solutions were developed in making this prototype
Smart Intercom System using TinyML deployed on Arduino33 BLE Sense: The following will be a touch-free solution using Computer Vision and a TinyML model to detect a person outside the door and conduct a bell ring without the person touching the bell.
TemperatureMonitoring system Using IoT and alert system : Amidst the pandemic, safety has become an important aspect. Hence, the Temperature Monitoring system utilizes IoT Thingspeak dashboard and detects people while entering and measures their temperature. This temperature is displayed to an IoT Dashboard for timely trends and data analysis. Upon abnormal temperature detection, alerts are generated upon which the person undergoes second inspection.
Touch-Free elevator system using Speech Recognition TinyML model on Arduino BLE 33 sense: We use lifts to go up or down in a building several times a day, and I always have a fear of touching contaminated switches which have been touched by other people commuting. Hence this speech recognition model will identify when a person wishes to go up or down and similarly will perform the action.
MaskModel Detection System based on TinyML and IoT monitoring system : This method will use a computer vision model deployed on Arduino BLE 33 sense to detect whether a person has worn a mask or not and similarly this data is sent to an IoT Dashboard for monitoring and imposing restrictions according to unsafe times.
Smart Queue monitoring and establishing system in a supermarket or a mall using TinyML, IoT and computer vision: This model will detect a person standing outside the supermarket and allow the entry of 50 people at a time in the supermarket. It will wait for another 15minutes to let the people inside complete their shopping and allow the next set of 50people into the mall again. This will be done using computer vision and TinyML deployed on an Arduino 33 BLE sense. This data will then be projected to an IoT Dashboard where real time data can be tracked.
Person Monitoring System in an Aisle in a Mall and contamination based Sanitization System: This solution Uses Person detection Algorithm deployed in an area in a Mall or a supermarket and if the person contamination in an area has passed the threshold, It self sanitizes the area with UV light. The Sanitization period and times is projected on an IoT Dashboard for the supermarket staff for analysis.”