Elavo - Space Apps Challenge

Back to Space Apps

Elavo| Purify the Air Supply

East Asia and Pacific

Awards & Nominations

Elavo has received the following awards and nominations. Way to go!

Best Use of Technology

The solution that exemplifies the most innovative use of technology.

Purify the Air Supply

Has your time spent indoors increased during the COVID-19 pandemic as a result of stay-at-home and shelter-in-place policies worldwide? Your challenge is to use the International Space Station (ISS) as inspiration and develop a system to monitor and/or purify indoor air. It is entirely up to you whether the system you design is able to be used on Earth (for example in homes, businesses, transportation, etc.) and/or in space.

Elavo | Breathe Easy

Summary

Travelling to and from work, or from floor to floor in a building, shouldn’t be a health hazard. COVID-19 has changed the way we think about the risks associated with these mundane activities. Elavo aims to keep your commute safe, by applying NASA technology to scrub the air of pathogens and automatically keep surfaces sterile in elevators, trains and buses. With Elavo, we can all breathe easy.

How We Addressed This Challenge

The Problem

The COVID-19 situation is fast-moving, and so must be our response. We need to get public transport operating at close to its previous capacity, but with greatly reduced risk of disease transmission. Currently, infection control is managed by social distancing measures and frequent chemical disinfection; however, social distancing is difficult to enforce, and chemical disinfection is labour-intensive and expensive.

Our solution

Elavo’s approach avoids these problems by integrating infection control directly into the building or vehicle. Our solution aims to reduce the potential for infection via contaminated air or surfaces in high-traffic confined spaces within the transport network.

In confined spaces, such as elevators or public transport systems, it is impossible to avoid breathing the exhaled air from other passengers. Many pathogens from this exhaled air will eventually settle on surfaces including buttons and armrests that will be inevitably touched by other passengers. Elavo is a solution to reduce air and surface borne contaminants. Elavo achieves infection control by combining advanced NASA photo-activated catalytic air purification technology with LED-based UVC photosterilisation of surfaces. Elavo’s solution can be retrofitted to existing elevators, trains and buses, and aims to substantially reduce the risk of infection through inhaled or contact-acquired pathogens.

In each type of enclosed space, the Elavo ventilation system draws return air from low mounted ducts, and provides processed air flowing down from high mounted ducts. The downward flow of air reduces the distance that exhaled droplets can travel. Elavo differs from existing ventilation systems used in these applications by the addition of NASA’s Airocide photoactivated catalyst technology. This technology is used to destroy pathogens and volatile organic compounds in the return air prior to processing through air conditioning and being returned to the space from the output ducts. To avoid infection through contaminated surfaces, high intensity LED-based UVC lighting is used to eradicate pathogens, both opportunistically and at periodic intervals.

For elevators, photosterilisation is activated at any time that the elevator is empty and the doors are closed; the elevator can also be programmed to enter a cleaning cycle when the last passengers exit at the end of its primary traffic flow direction (e.g. at the top floor in the morning). The hue of the elevator lighting is changed to red, and voice indicators are used at this time to indicate that all passengers should leave and no-one should attempt to enter until sterilisation is complete. The lift remains immobile until the cleaning cycle is performed.

In trains, buses or ferries, photosterilisation is performed automatically at the terminus; again, lighting and voice cues are used to ensure that all passengers exit the vehicle during the cleaning cycle.

Thermal cameras are integrated into the system, together with automatic acoustic detection of sneezes and coughs, to increase the rate of airflow if potentially infectious individuals are detected. Elevators in secured buildings (such as hotels and apartment complexes) which require authentication prior to departure can also correlate sneezes/coughs and elevated temperature detection into automatic contact tracing systems, by recording passengers who travelled in the proximity of potentially infectious individuals. Tap-on/tap-off systems in public transport networks can also be used to enhance contact tracing in some cases (for example, on buses).

How We Developed This Project

Inspiration

Capacity on public transport has been slashed by ⅔ to maintain social distancing, but many people have no alternative for transport. In some cases, employees are forced to use crowded elevators to access workplaces, and there is no way to know whether other users are sick or have been practicing good hygiene. Elavo explored the data using R with the aim of evaluating the assumption that higher public transit use was correlated with higher infection rates - and we found that this was indeed the case. With this knowledge, we wanted to see what we could do to make using these services safer, without adding huge maintenance or operational costs.

Approach

Our approach centred around identification of technologies which are commercially available and ready to deploy in a short timeframe with minimal disruption to current operation of transport and mobility systems. We also looked for potential opportunities for disease identification contact tracing which would be useful not only for COVID-19, but also for outbreaks of other diseases in the future. To validate our hypothesis, we performed a case-study approach using New York as an example to justify our projects specifications and capabilities. We used a combination of satellite EARTHDATA supplied from NASA to identify current trends of one major pollutant outdoor air quality; nitrogen dioxide, with statistical software (R Studio program) and geostatistical analysis (ArcGIS Desktop) to highlight the correlations between public transport connectivity (specifically train and subway stations/underground lines), air quality and COVID-19 case distribution.

Data-driven approach case study: New York

New York and its surrounding suburbs was and still is a major epicentre of the deadly COVID-19 epidemic in the United States. A recent study has highlighted its reliance on public transport may be a driving force behind the outbreak.

To illustrate the intensity of public transit connectivity and number of COVID cases in New York, maps were generated using open access repositories from ArcGIS Online. Datasets that were converted into feature layers included the; Intermodal Passenger Connectivity Database; nationwide data table of 7,000 rail, air, bus and ferry passenger transportation terminals; most up-to-date COVID-19 cases for the United States sourced from WHO, CDC, NHC and state government health departments collated at John Hopkins University* (*data is updated every hour, so maps are fixed to current statistics as of Saturday 30th May).

https://drive.google.com/open?id=1ISNFPUas1kamfQsv49NSaPWpWvnR6WBC

Key observation: New York has an extensive public transport network, compared with Los Angeles and Boston. New York provides above the average 2 modes of connectivity, that contributes to the outdoor air quality and hypothetically be a contributing factor to the transmission of the COVID-19 virus.

To investigate the spatial and temporal variation of outdoor air quality, satellite data was extracted,in order to extrapolate the linkage between pollution from the activity of New York and incidence of COVID cases. Nitrogen dioxide is a toxic component as a result of human activity, but also causes inflammatory responses in the airways. Chronic respiratory stress is likely to limit the body's ability to defend itself for new emerging viruses, thus NO2 levels were investigated and compared before and after the initiation of the citywide lockdown. For the NO2 concentration in the troposphere (from surface up to ~10 km), satellite imagery was obtained from the Sentinel-5 Precursor and processed in GIOVANNI NASA EARTHDATA software; NO2 Tropospheric Column (30% Cloud Screened). Comparisons were made between the first two weeks prior to the first known COVID-19 (24th February-9th March) and the lockdown and the week of lockdown initiation (2nd-16th March). Quantitative data was also gathered from GIOVANNI by presenting histograms of accumulative daily NO2 levels over the same time period.

https://drive.google.com/open?id=1lMHKX2SXS13WMsf4nak5Xnu72FNe8ej-

Key observation: From mapping the tropospheric NO2 levels over New York (to a maximum resolution) it reveals a major ‘hotspot’ of high concentrations (100-175 micromol/m2) prior to the lockdown, especially stretching from East Village in the South to Upper Manhattan/Bronx. During lockdown, levels significantly decreased between 50 and 90 micromol/m2. However, the persistence signals the air quality of New York may intertwine with emissions from public transport to contribute to higher risk of respiratory diseases. Research has continually shown neighbourhoods including the Bronx, Manhattan and Brooklyn continually struggles with air quality issues and high rates of asthma. It is not surprising that these areas also have the highest number of COVID-19 cases (per 100,000 people). Regardless, this illustrates even if with less people out in the streets or using public transport, contaminated air is still present in the atmosphere. Continual exposure could subset inflammatory responses in the lungs, increasing the possibility of infections through the use of public transport where air is often recycled from the outdoors. This initiated further investigation into the relationship between high-traffic public transport and current cases of COVID-19.

ArcGIS and built-in geostatistical analysis tools (GeoStatistical Wizard, Semivariogram Covariance Cloud) were utilised to calculate a 2D cross correlation, using the covariance of COVID cases in New York in relation to the location of public transport transit stops.

A Kriging tool was used to generate a layer that uses probability models that fits best to a scattered set of point values (distance between COVID cases) in two-dimensional space. It assumes a uniform pattern of point value distributions (equal distribution of COVID cases), so a semivariogram was created.

Ordinary Kriging was used because the constant increase in the mean number of current COVID cases across the spatial scale of cities is unknown.

A cross covariance tool was then used to investigate the cross-correlation between two different datasets of interest.

Cross correlations were performed in ArcGIS Map 10.8.

Datasets consisted of:

CovidCaseby_NYCZip = Number of individual COVID cases in New York by zip code

NewYorkCitySubway= Stations along the subway/train lines; Brooklyn, Manhattan, Queens, Bronx, Staten Island

Kriging Prediction Map was analysed based on the semiogram between the a) the number of COVID cases in NY as of 27th of May and b) the subway and train stations/lines.

https://drive.google.com/open?id=1J1FUnidRcQpCZChbITwYbI5VFgWVATDU

Through the Kriging model (identify 'hotspots' of COVID cases),, it shows to be asymmetric. The darker area shows the highest correlation between subway stations and lines occurs and a higher density of confirmed cases of COVID. Two key areas of focus with a higher number of cases are located towards the north and east of the city.

A crossvariance cloud was also performed to graph how semivariance changes, as the distance between observations (number of COVID cases) changes. This also takes into account the location and proximity to train and subway stations/lines. The graph shows the relationship between a point at an origin with another point, according to its distance in the x and y direction (the point chosen to be at the origin is the one that will keep the vector in the first or second quadrant).

https://drive.google.com/open?id=1r6DiqN9IYZ1K0X7c-FzebUaVma7XtAfQ

From directing the point of focus north of central New York, it clearly shows the linkage between a potential hotspot of COVID cases and a high density of train/subway stations. Higher values may occur because the hotspot area is correlated with a network of train and subway stations/lines.

Learning from these previous observations, statistical analysis using R Studio programming was performed to assess the correlation of two numerical vectors (number of daily cases in New York versus percentage change in the number of people using public transit systems) along a time-series in order to find the maximum correlation coefficient.

https://drive.google.com/open?id=1lzFKwC6jI0ejjMV_x5c0Hrw1dVcsNJqS

This measures how long it takes for an effect to propagate from one variable to the other, in this case the effect; number of daily cases in New York and measured variable would be the percentage change in the number of people using public transit systems). Data was supplied from Moovit that recorded the impact of COVID-19 on public transport usage (usage of the previous 7 days in New York was compared to a typical week before the outbreak began (the week prior to January 15th): https://moovitapp.com/insights/en-gb/Moovit_Insights_Public_Transport_Index-121 and from New York City Health: https://www1.nyc.gov/site/doh/covid/covid-19-data.page

https://drive.google.com/open?id=15SJ5l5wO0XL2NLoqBHrtHT9nGNRc1s-x

Key observations: As shown in the graph, the maximum cross-correlation is approximately 0.6, which signifies a fairly strong correlation. This could infer an increase in the number of COVID cases likely influenced people’s perceptions of not using public transport networks. Keeping in mind, prior to lockdown there was a two week period where people could still utilise transportation. However, it is interesting to note this maximum correlation happens at -1 and +3 periods.

The correlations in this region are positive, indicating an increase in the number of daily cases in New York is likely to lead to a greater decrease in the number of citizens using public transit stations.

Another key observation is the trough between +12 and +15 (this is the period when the lockdown was initiated in the city). The correlations in this region are negative, indicating that a large decrease in the percentage of people using public transit stations likely contributed to a decreased number of daily covid cases. Though this is all hypothetical and prediction based, it illustrates the impact of a lockdown on people’s mobility actions but also how it could predict the ongoing halt of public transport systems.

By only focusing on a filtered dataset, we could perform an analysis using the starting date of city lockdown (13th of March 2020) until the 27th of May 2020, to see the gradual impact of the lockdown measures.

Surprisingly, the maximum efficiency is higher and more negative (close to -0.6).

As it is seen, at t=-1 and t=0, the correlation coefficient t=-1 is 0.449, which means the number of cases in NY following the day of lockdown is correlated positively with the percentage change of people using public transit stations in the city, the following day. As lockdown progressed, the number of people using transit stations decreased.

However between +10 and +15, the correlation lag is -0.6 (greater than a significant level of p=0.5), so the correlation could be identified as significant.

It is possible to infer the number of COVID cases in New York significantly decreases as there is a decrease in the percentage of people utilising transit stations(large percentage drop in public transport utilisation).

Project Demo

https://www.youtube.com/watch?v=p0NoPin5lyc

Use case 1: Building elevator

Elevator in sterilisation mode and operation mode

Elevator controls

During a one-minute trip in a 6-person lift, allowing for body volume, available air in the lift and average human per-minute ventilation rate, about 2% of the air you breathe is someone else’s exhaled breath. In a large apartment building, you have no feasible alternative means to reach your floor. Viruses and bacteria can easily spread through exhaled particles and contaminated surfaces.

With Elavo, the risk of both airborne and contact infection can be drastically reduced. Consider the scenario below:

Several users wait for the elevator on the ground floor. The freshly sterilised elevator arrives; the doors open. Each person taps on, identifying themselves, before each selects their floor. A thermal camera checks the temperature of each person. A user is identified as having a temperature of 38.5 degrees; a warning is shown on the display panel. The ventilation rate is increased in response, and the identity of all users who have tapped on are logged for contact tracing if required. The elevator departs, depositing users at their respective floors. Upon reaching the top destination floor, the elevator lighting turns red, and a voice notification declares that the elevator is about to be sterlised. One person waiting to travel to the ground floor pauses and waits for the cycle to complete (for approximately two minutes). The doors close, and the UVA lighting system is activated while the ventilation system continues to clean the air of potential pathogens. Finally, after two minutes, the lights and airflow rate return to normal, the doors open and the waiting user is welcomed into the elevator for their descent.

Later, the person with the elevated temperature reports for a COVID-19 test, which returns positive. The test centre uses data recorded by the Elavo system to contact-trace the other users of the elevator and begin the process of testing.

Use case 2: New York subway

A NYC subway carriage has a rated capacity of 258 people, although in reality, this limit is often exceeded. The volume of air in each carriage is approximately 80000 L; a train at rated capacity will result in the production of in excess of 2000 L of exhaled air per minute, making up 2.5% of the air volume in the carriage. In between two stops, with an interval of 4 minutes, 10% of the carriage air volume is exhaled breath. Longer journeys naturally result in a higher proportion of exhaled air in the atmosphere, with some mixing with fresh air occurring at each stop.

The use case for Elavo for a subway train would be as follows:

The train departs from the subway yard, having been sterilised automatically before the driver boards prior to its scheduled departure time. As it arrives at the first stop, users start to board for their morning commute to work. The train rapidly fills to capacity, with many users standing. The ventilation system runs continuously, cleaning the air as it circulates throughout the car. As each user enters, they are scanned for elevated temperature. The acoustic sneeze/cough detector is triggered by an elderly passenger; as for the elevator use case, the ventilation system temporarily increases the airflow rate in that car. When the train arrives at its final stop, the last passengers leave, and the driver waits outside for the sterilisation process to complete, before starting on the return journey.