As alluded to by Dr Sheng, the supply chain has its part to play in minimising damaging air emissions, while simultaneously improving health and wellbeing.
Mixing air with artificial intelligence
Poor air quality generates significant debate with people questioning who is to blame, as well as what are the causes and effects. Studies exist(6,9), yet they potentially lack two core criteria: the first is a granular approach to obtaining significant data readings of the key particulate matter; the second is obtaining this data over a continuous period of time to be able to determine insights. This frequently results in short-term surveys which lack objectivity and detailed evidence.
To reconcile this, a plausible first step could be for a community to build a basic system on commodity sensors, transmitting this data over a shared network (typically WiFi) to free websites that can display these results (these are already operating today). The systems are typically based on open-source technologies, so free to use to an extent, and a great first step in identifying the problem, albeit limited in functionality.
Commercialising this approach provides the next stage. By deploying a concentration of sensors that are able to collate significant data, added with real-time reporting, provides richer data that results in objective facts. This approach also benefits from the application of algorithms with other data sets, such as weather, industry and transport, allowing us to predict the areas that need the most attention and prioritise these first.
Dr Dawid Hanak, a senior lecturer in Energy and Process Engineering at Cranfield School of Water, Energy and Environment advises: "It’s not just particulate matter that we need to be concerned about. It’s also an increasing CO2 level that significantly affects our health and wellbeing. Despite our efforts to curb emissions, mostly via investment in and incentives for renewable energy, the atmospheric CO2 concentration is still increasing rapidly(11). It’s alarming that we are starting to see the dire consequences of global warming, such as the ice caps melting, the increase of sea levels causing flooding of the coastal areas, the decline of lakes and drastic weather changes(12).
To meet the UK’s net-zero commitment by 2050 as outlined by the government, the industry must deploy carbon capture, utilisation and storage along with renewable energy to achieve the desired reductions in CO2 emissions from power and industrial sectors in this current timeline(13). This technology has already been applied in an industrial setting for gas purification and hydrogen production since the 1930s(14). However, it’s wider deployment to other industries, especially for generating electricity with low CO2 emissions has been delayed. This is due to high costs associated with mature off-the-shelf technologies, risks with new significant bespoke products and insufficient investment incentives. Emerging technologies show a promise of not only lower equipment and operating costs, but primarily of innovative business models that enable industries to engage in the circular economy, reaching their corporate sustainability commitments and maximising their profit in the long term. Further R&D activities can be significantly enhanced with the new AI capabilities that can support the process development cycle and simultaneously optimise multiple possible business models."
As Dr Hanak points out, the technology to tackle the problem exists. With data, artificial intelligence and machine learning we can pin-point key areas that will have a maximum impact, underpinned by an economic model, and therefore promote health and wellbeing as well as economic progress.
Getting to the data
Without doubt this problem will not disappear or improve overnight. It also requires multiple approaches to improve the situation. Community-led initiatives to record air quality are the first step towards a granular approach, which could then benefit from AI capabilities designed to provide insights and support plans to tackle this problem. Enabling this mass data collection requires a secure, robust network, which is where IoT (the Internet of Things) comes in. It provides a secure, manageable eco-system to allow AI algorithms to provide insights needed to address this problem.
The low cost of sensors and availability of a mix of communication systems provide the base infrastructure and data collation ability. It’s relatively simple to build on this and deploy a granular, resilient and secure IoT eco-system to provide the insights we need so we can work out the approaches we need to take.
The problem we have isn’t a technical challenge; it’s an organisational one across industry, government, and most importantly, communities.