A while back, I wrote a post about testing an earlier version of the Pollution Guardian prototype in the University of Surrey’s Envilution® laboratory test chamber.

A key point of using the chamber was to be able to objectively assess how well individual particulate matter and gas sensors performed, compared to high quality laboratory measuring equipment, in the controlled conditions of the lab test chamber.

Pollution Guardian prototypes installed inside the Envilution® chamber at the GCARE’s Air Quality Laboratory

Any issues found with sensors in the lab, are likely to be more significant in real world testing. In fact, the Envilution® chamber has helped us triage some significant issues with gas sensors over the course of the project.

The collaboration with the University has extended beyond the lab into the on-road field testing activities. On multiple occasions we have tested a car loaded up with Pollution Guardian prototypes and high grade laboratory test instruments. This was in order to compare the performance of prototypes in real life conditions i.e. with rapidly changing levels of air pollutants combined with environment changes to the car interior via the action of the car air conditioning system.

High quality test gear installed in-car alongside Pollution Guardian prototypes

In the video below, Professor Prashant Kumar, Founding Director of the University of Surrey’s Global Centre for Clean Air Research (GCARE) summarises the status of the Pollution Guardian work and his team member, Dr. Hamid Omidvarborna, tells us about the Envilution® test chamber.

Thursday 17^th June has been nominated as “clean air day” in the UK so a good time for an update on the Pollution Guardian project.

As mentioned in earlier blogs, the Pollution Guardian is our project to develop a low cost solution for monitoring and alerting road drivers to areas of high air pollution and so enable them to take counter measures. We have been working on this project together with the University of Surrey and it is supported by an Innovate UK grant to prove the concept and further develop prototypes.

We have now come to the conclusion of the current phase of work which has involved:

• Construction and characterisation of 20-30 prototypes with an updated electronic and mechanical architecture
• MVP (minimum viable product) development of a mobile app geared to consumer test
• Cloud back-end development to support data capture and visualisation on the mobile app
• Collaboration with a UK facility management business to run a solution trial in Q1/2021
• Analysis of pollution data and user experience feedback to guide the next steps
• Creation of short films to help us explain the project (see here)

A couple of credits here to the people who have been helping us achieve some of the above:

• Peter Roll of Roll Design Ltd for mechanical engineering work on the prototypes
• Mark Mason of Design Thinking Ltd for industrial design and user experience (UX) work
• David Savva of Alchemy Street Ltd for film direction and creation

We will be creating a few more blogs and releasing some additional films covering the more recent work in the coming weeks, so watch this space.

Introduction

As we arrive in the final quarter of our Innovate UK sponsored project and with an ongoing Covid pandemic, it is worth us asking the question why are we bothering with the Pollution Guardian? Let’s start with the facts:

• Poor air quality contributes to reduced life expectancy and a significant number of deaths every year. (ref p.110)
• Motor vehicle emissions are significant contributors (ref p.36) to poor air quality, especially in urban areas.
• Motor vehicle users face the highest level of exposure to air pollutants compared to other modes of transport.
• Significant numbers of people identify poor air quality as a trigger for their respiratory issues.

Thus, we have a significant health and air quality issue, made worse by vehicle use, but which is potentially most harmful to the vehicle users themselves. The obvious solution would be to get more people to drive less or drive more zero emissions vehicles but:

• Changing habits and lifestyle is tricky to pull off.
  • Whilst we have achieved remarkable things during the pandemic, are we motivated enough to address the silent killer of poor air quality?
  • Unfortunately, the evidence is pointing otherwise as ride as a service and online to home delivery services can push up road congestion and air pollution
• What about electric vehicles?
  • The UK government’s stated aim is to ban sales of new combustion engine vehicles starting from 2030/2032/2035 depending on whom you choose to believe.
  • This is a potentially worthy objective but there are big challenges: charging points, range anxiety and the affordability of electric vehicles.

So, what can we do about it now?

The Idea

Well, whilst in a car as driver or passenger, you can control your own ventilation by rolling up the windows and using the air recirculation button. This has been shown to have a big impact on exposure to poor air quality .  We took the view that it ought to be possible to make an affordable sensor to warn car occupants of poor AQ and so take countermeasures, if we could incorporate existing technology items, smartphones in fact, into the solution. Thus the pollution guardian project was born.

Now we are some way down the line since starting out, how can we assess if what we have been developing could work for people in practice? Our answer: to organise a real life trial incorporating real users, not associated with the project, and let them use the device for a few weeks and listen to their feedback.

• Are they more conscious of their air quality whilst driving?
• Has the use of the Pollution Guardian influenced their behaviour at all?
• What do they like/dislike about the solution?
• Can we show any differences over time from the data that we gather?

Thus we are engaged now in putting together a real life trial and the next blogs will focus on the various aspects around this.

Much has been made over the reduction in air pollution in 2020 during the Coronavirus pandemic. Certainly, the first period of lockdown saw much reduced commuting activity as people were asked to stay at home or work from home and it was certainly my perception that the volume of traffic was less outside of the main commuting times.

Deserted A3 carriageway near Guildford

However, this description and picture above are not accurate descriptions of 2020. Driving around West and South West London in August showed that there was indeed plenty of traffic and plenty of air pollution around.

Busy Great West Road in West London in August 2020

As this blog is motivated by our Pollution Guardian project, you might be interested to see what we observed on the route around West London. Below, some estimates of the Nitrogen Dioxide gas concentration taken en-route from inside the test vehicle over a 10-15 minute period.

Pollution Guardian prototype NO2 gas level estimates (ppb) against time (s)

Note that whilst the Pollution Guardian prototyping work is still ongoing, the absolute numbers in the graph above were our estimates based on the individual prototype sensor outputs and algorithms at the time. They are however valid to look at as a high quality, lab grade, NO2 gas analyser on board our vehicle, also reported similar or higher gas concentration figures over the same time period.

What does this all tell us?

• Despite the pandemic, traffic and pollution levels over the summer period in city/urban areas were not so much below the “business as usual” levels.
  • Indeed, the BBC reports that even in the midst of the January 2021 lockdown, motor vehicle traffic is already at 50% of normal levels.
  • This suggests traffic and its associated air pollution will be returning to normal in 2021 when the lockdown eases.
• Air pollution levels can vary a lot from lower to higher levels of concern over relatively short distances i.e. hotspots are discrete spots.
• The Pollution Guardian looks capable of following these fast changes in air quality and can be a useful tool to bring awareness and raise alerts on worsening conditions.

As part of the Pollution Guardian project, our collaborators, the University of Surrey Global Centre for Clean Air Research (GCARE) planned to develop an air pollution test chamber for affordable sensor devices to be evaluated against high quality laboratory grade test measuring instruments.

The objective for the chamber was to be able to set up stable, predictable air pollution and environmental conditions to enable the replication of field conditions from ANYWHERE in the world.

In its first iteration, the chamber, subsequently branded “Envilution”, was capable of controlling:

• Chamber air temperature
• Humidity, including pre-filtering to create “zero air” as a baseline (dried and filtered air)
• Particulate Matter (PM) levels, with options to create particulates from a variety of sources
• Nitrogen Dioxide, NO₂, gas concentration

By the end of the second quarter of the project, the chamber was ready to start testing with our Pollution Guardian prototypes.

Figure 1: Envilution air pollution test chamber created by the University of Surrey

The chamber conditions:  temperature, relative humidity, particulate matter level and NO2 concentration were sampled from a port in the chamber using the same high quality instrumentation mentioned in the previous blog.

In our testing, we planned on exploring the behaviour of the prototype sensor units against changes in the levels of temperature, humidity, particulate matter and NO2 gas concentrations. Whilst some of these tests were more straightforward to set up, others turned out to be more tricky e.g. humidity testing.

Mostly when weather websites predict humidity levels, they actually refer to the “relative humidity”. This is a measure between 0-100% of how much humidity the air is capable of carrying, relative to the air temperature and pressure. As you get toward 100%, you can get a lot of condensation forming on surfaces and in warm conditions struggle to keep cool by sweating.

In our lab tests, we aimed to see how our sensors responded overall to the relative humidity in the range of 10-85%. The initial approach was to use an aquarium to generate the humidity and mix this with “zero air” (cleaned and dried) to achieve a target relative humidity.

Figure 2: Zero air combining with water vapour from the aquarium

As we discovered, it was quite hard when starting off with zero air to turn this into high humidity air without also having condensation issues within the ducting leading to the test chamber. The GCARE team were however quite resourceful when coming to practical techniques, and we managed to explore the other parts of the humidity range using a combination of a nebuliser to create small water vapour particles and the higher end range of humidity through rearrangement of the sets of connections into the chamber.

I will tell a bit more about the performance of our sensors in the chamber in subsequent blogs whilst briefly mentioning the results from our temperature sensor during a temperature test.

Figure 3: CP1 prototype temperature responses during a lab heating/cooling experiment

The CP1 prototype units’ temperature sensor responses were generally well correlated with one another, but slightly higher than the chamber temperature recording. This was attributable to the internal heat dissipation of the prototype, when in operational mode.

One of the first activities together with the University of Surrey’ Global Centre for Clean Air Research, led by Professor Prashant Kumar, was to join a live field campaign to measure the air pollution next to Stoke Road in Guildford, Surrey. The University of Surrey’s campaign measured roadside (pavement) air pollution using both high-end specialised air pollution instrumentation along with more affordable devices.

Stoke Road can be a busy road at peak times and is closely connected to the A3 slip road for London and the South East as well as the direct connecting route to the nearby town of Woking.

Figure 1: Stoke Road during a morning rush hour.

The setup involved measuring the air pollution at a height of approximately 1.5meters above the pavement, a comparable height for pedestrians breathing in roadside air. Thus the need to brush off some old woodwork skills and create a container and support platform for our units.

Figure 2: “bird table” box for holding pollution guardian units

Why measure pollution on Stoke Road? There is a children’s playground right next to Stoke Road, so monitoring pollution levels there is of real interest and also the University has a co-located air pollution measurement station, using sensors from the iSCAPE project. With the support from the Guildford Borough Council, this site is being continuously used by GCARE team as a part of their iSCAPE’s Guildford Living Lab activities.

Now whilst we are designing the Pollution Guardian for measurements in the car, it was of some interest to us to obtain an early view of how our own “CP1” version prototypes performed compared to the specialised air pollution equipment made by GRIMM and more expensive particulate material testers such as Dylos devices.

Figure 3: Roadside setup next to Stoke Road

What were the learnings?

Good results are possible for fine particulate matter (PM_2.5) measurement:

• After aligning measurements in time, we are able to obtain a correlation of 0.8 between Pollution Guardian units and the high-end particle spectrometer.

However, NO₂ pollutant gas measurement is difficult in the field:

• Pollution Guardian NO₂ sensor readings were well correlated with one another but..
• Even after warm-up, we can only obtain 0.3 correlation with a reference measuring unit
  • Some NO₂ spikes detected by Pollution Guardian units were ignored by the reference NO₂ detector and vice-versa.

• One effect we could feel and could see in the “spikiness” of our NO₂ readings was the impact of a Northerly wind blowing along the road.
  • We would go on to look at how to better protect our units against wind gusts, whilst not compromising our key use case of the car environment.

So, whilst the reasons for lower correlation on NO₂ gas readings were not yet clear, this was something we were obviously keen to explore in the planned lab testing with the University of Surrey’s GCARE Air Quality Lab where we would have better control over the environmental conditions.