Sustainable Finance Live 2023: Democratising geospatial data and informing AI usage

Ordnance Survey has over 230 surveyors in the field that provide 20,000 updates a day and uses drone technology and satellite data. Lyndsay highlighted that satellite data allows the data to be kept fresh and up to date, and can be help inform the attribution of data using multi-spectral characteristics. She shared how data science is able to detect water pollution through the use of Smart Biospheres developed using geospatial technology:

“We spawned a project thinking about what are the some of the key pain points that going on at the moment, and what is possible, and how can we provide solutions? We worked with CGI to look at what we could do with data. We did trials and created models to look at whether or not we could detect sewage from space. Short answer is it's really hard. When you start putting together loads of data sources, you can start trading tools to help you predict when these events are going to happen. One of the models created was relatively good, and now we are testing it in the real world with the smart buyers here in North Devon, and they have over 50 sensors deployed in one of the river catchments so we can test whether the predictive algorithms are actually predicting the pollution event accurately enough. It's no good having a black box technology where you are not sure what the inputs are and what the outputs are.”

She emphasised the significance of data collected by satellites, as they can assist in facing unprecedented change when it comes to climate change using historic data which can help understanding and monitoring of climate change as it progresses.

Open data can map thermal signatures that can chart where cooling needs to take place and where the heat levels are unsafe, informing urban planning designs for how to create spaces where natural cooling can be implemented and health services can be placed. The combination of data and intelligence can inform planning, visualise where change needs to take place, and form nature-based solutions to see what does and doesn’t work in the field, Lyndsay explained.

She continued that Ordnance Survey is prioritising data democratisation by using ChatGPT to allow people to monitor how their local areas are being impacted by climate change. Through their applications, people will be able to read thermal signatures, view flood zones, and learn more about the status of their environment.

She stated: “I think we can really democratise information to make it useful. People don't have to be specialists in geospatial information to use the data appropriately and meet their needs. We can make sure to control that input, make sure it's quality, that in the black box scenario you can lift the lid and see what's inside so people can trust what's going on.”

Alongside her explanation on the benefits and numerous uses of AI and data, Lyndsay warned on misuse of AI technology and data models that can share misinformation, develop bias, and distort the data output. People are able to cease emissions when satellites are tracking their data, and so they can try to game the system and manipulate the data collected to make it seem better than it is.

Lyndsay concluded that what Ordnance is doing as present as taking knowledge that has been accrued and applying it globally, collecting data for a purpose and data that delivers value. Through Ordnance’s technology, they would be able to find emissions that people are trying to hide based on location.

Following the keynote session, Darshna Shah, director of innovation, Elastacloud, moderated a panel on the role of behavioural science and the potential winners and losers when cities change to meet sustainable finance goals.

Joining Shah on stage was Guillaume Levannier, sustainable finance board member and asset manager, Lombard Odier Investment Managers; Walid Al Saqqaff, founder, ReBalance Earth; Jannika Aalto, program manager, sustainable circular cities and built environment, Green Digital Finance Alliance; and Adrian Sargent, CEO, Castle Community Bank.

Leveraging her background in neuroscience, Shah provided a fresh perspective on approaching sustainability and presenting the subject through from a scientist’s perspective. “We’re bombarded with a lot of data, it can seem overwhelming, and to form habitual change, we need to engage our prefrontal cortex, apply a lot of attention, decision making and to some degree, emotional regulation.

“And with that, we need to repeat behaviours before they can become habits. We have a lot of contextual stresses on us, whether that’s pricing and products or social norms, that may not always lead to the value and behavioural link that we expect,” Shah added. She went on to reference datasets provided by the World Resources Institute and highlighted three key studies into the Covid travel ban, community energy use reports, and menu language – which revealed sustainable benefits.

These include the realisation that work can be conducted efficiently remotely, awareness of neighbours’ energy consumption can help reduce usage, and adjusting how vegetarian food is described can reduce the number of meat dishes ordered. With this, Shah proved that “small incremental changes could lead to more widespread sustainable behaviour.”

But how can financial institutions plan for this, and what are the most important behavioural considerations? Walid Al Saqqaff, founder, ReBalance Earth, highlighted that nature must be considered in a very different way to carbon. “Whenever you want to prove biodiversity, you have to think in terms of decades, not centuries, and to do that, you need the buy in of the local communities. You need to understand what their existing behaviours are, why they established in that manner, and how you can move towards something that is more sustainable?”

Al Saqqaff also mentioned that aligning to the UN SDGs is an effective way of filtering through insights and understanding what the community is prioritising. “That can act as your lighthouse,” he added. A second consideration is to look into areas within cities that have been deprived for historical reasons, investment has dried up, and as a result, employment has suffered. By understanding the challenges that communities are enduring, the solutions to these issues can be extracted from data, as explored in the previous session.

While data can be gained from satellites or drones, Al Saqqaff explained that one of the greatest sources of information is in fact through people, “citizen science”, and creating that “feedback loop to correct AI models to apply a policy of equity and fairness. But are we all on board? Sargeant stated that all those attending Sustainable Finance Live are “motivated and want to see change, but it's a question of how do you then get that change through and how does that legislature or regulation come through?

“We need the support of the cities to allow the framework to happen, but then, we also need the finance to flow through. Both councils and governments need money to support it, which can come through private investment,” – although this is occasionally difficult to garner for certain types of communities and companies. All panelists agreed that partnerships are the way forward, particularly to ensure that frameworks such as these can be funded.

Jannika Aalto, program manager, sustainable circular cities and built environment, Green Digital Finance Alliance, drilled down into this and said that although the term “cocreation” has become a buzzword recently, but “there is real value in meaningful engagement early on with local communities.”

Guillaume Levannier, sustainable finance board member and asset manager, Lombard Odier Investment Managers, provided a concluding comment on the use of data and said that when working with small to medium sized businesses, in addition to engaging with the communities, banks must also focus on the “risk of buy ins into ideas, and the knowledge about the risk.”