Artificial intelligence (AI) – much faster, better and lower-cost client solutions

02 Aug Artificial intelligence (AI) – much faster, better and lower-cost client solutions

We, and the world in general, hold huge amounts of data. I would like to examine how we can use this information with AI to revolutionise project design and planning – and also look at how mathematic algorithms are helping to solve many difficult environmental problems in the field.

I should perhaps start with a definition. Artificial intelligence is the science of making machines that can think like humans. As a result, AI is not only able to process massive data sets, but also recognise trends and make smart decisions in real-time.

The environmental potential is massive. AI can ‘understand’ complex flooding patterns and predict major catastrophes. On a local scale, it can improve energy-efficiency, reduce both waste and raw material use, and identify supply chain carbon emissions that must be eliminated to reach ‘net-zero’ by 2050.

Replacing humans? Not yet

When it comes to risk calculations, AI can be used to enhance or replace climate models that would have involved enormous person-power – and a lot of concentration – in the past. Despite this, it is not actually a brand new technology and has its roots in the work of computer pioneer Alan Turing in the 1950s. But its day has come.

I want to look closely at how AI helps us to best manage, retrieve and interrogate client data as a specialist environmental consultancy. However, later I have also summarised some of the major benefits AI is bringing to global environmental challenges.

The power of data

As environmental consultants and data guardians, Enzygo (www.enzygo.com) has always had to securely store, sift, screen, search and filter masses of valuable information collected and sent to us by stakeholders and government organisations. Add to this historic material, site-specific facts and figures, and our own empirical data recorded on-site, analysed in labs, and from desk studies.

Most organisations are now in the same boat and hold bulk information that is often difficult to retrieve and takes up valuable server capacity. And much of it is never used!

With AI, many hours of work can be replaced at the click of a button – with additional value added. Searches for illusive project details on the edge of one’s memory that can perhaps only be hunted down through some local quirk about, say, planning, drainage, or noise, are now easy with AI.

Fewer hours mean lower costs. The easy sharing of data also means that we can draw on wider raw information sources before sharing processed data back with co-consultants and third-parties. The result is usually more robust and comprehensive end solutions. What is not to like?

Beyond data

However, that is not the end of the story. The processing ability of AI as it continues to be developed is enormous and has given us – and will continue to give us – enhanced modelling tools. These in turn are producing efficient and very practical algorithmic solutions.

While there are still well-publicised fears about AI’s future ability to destroy mankind, just as they are learning to cope with social media, businesses and consultants must be able to navigate around AI’s consequences in society.

Our philosophical approach at Enzygo is that, as a team, we must invest in moving ahead of the field now, and then learn continuously how to apply this booming technology to everyone’s advantage.

Enzygo

Enzygo is an independent multi-disciplinary environmental design and planning consultancy delivering creative, joined-up and cost-effective solutions that maximise the potential of individual development sites.

Our goal is to be a one-stop-service-shop producing detailed proposals that anticipate what local planning officials and elected council members require to make positive decisions.

To do this, we use conventional and emerging techniques and technologies. This is a natural fit with AI, and a logical extension of what our expert teams generate in terms of quality and content to gain rapid planning consent with minimal conditions attached.

Climate modelling

Better modelling can help us on a small and large scale. We use Government recommended climate change allowances in our work to understand rainfall intensity in drainage models, river flood levels and coastal flood levels.

AI will no doubt play an important part in future IPCC reporting, which will then influence future Environment Agency recommendations. The upshot is that advanced modelling should give us valuable new insights into climate risk assessment, disaster management, and resilience planning.

AI removes barriers

However, there are obstacles. The normal pre-application route can be blocked by local planning authority time, personnel and knowledge constraints. Developers/clients may also have contractual obligations to submit potentially successful applications regardless of the depth of supporting information needed and the guidance available.

At a minimum, planning proposal information needs to be validated. It is also important to be able to predict with some accuracy what issues are likely to come out in the wash. Experience and expertise are vital here. But AI could be a significant game-changer.

Ending difficult choices

This explains the current planning dilemma I believe.

It is tempting to wade in with early applications to make up for the fact that the system is sluggish. Difficult decisions may need to be taken, on the one hand, between working with consultancies that deliver on time and budget, and on the other, those who quote acceptable costs simply to get work signed off in a rush by investors, landowners, and strategic land partners.

To complicate the picture, clients may also be managing several third-parties simultaneously – some more communicative and trustworthy than others. The result can be a compromise. But with AI’s help, I think there is a better way.

AI to the rescue?

With its huge access to information, AI is able to identify very rapidly the individual development-specific data reporting requirements individual planning authority expect to be given.

More than this, we can now call up historic records of, say, planning appeals won or lost according to their type and scale, the percentage chances of delegated decisions by planning officers, and the general outcomes of certain planning committees on large residential schemes in specific locations.

With all this data at our fingertips, we simply need to know what questions to ask the AI system. When several consultants work together, they should be able to feed joint data into a Google-style supercomputer and run the same exercise. The potential potency is exciting!

No, it is not cheating!

Applicants must still comply on all matters. But by anticipating and responding to issues in advance, they should be able to save time, money and effort for everyone involved.

To summarise, AI will allow us to forecast accurately the chances of success, absolutely essential assessment requirements, and the scale of development if phasing is involved – plus firm prices with no awkward follow-up fees 12 months later!

Problems and solutions

But AI is not all beer and skittles. A significant concern is the amount of energy needed to train and operate AI algorithms. Unless this is from renewable sources, greenhouse gas emissions could rise.

However, data is increasingly showing us, firstly, that we need to act decisively to avoid the worst impacts of climate change, and, secondly, that it has a role in the solution. One key benefit is that AI can optimise energy use and reduce waste.

Machine learning algorithms are designed for computers to ‘learn’ from data, identify patterns, make predictions, or perform tasks without explicit programming. As a result, they can analyse smart grid data to optimise energy use while at the same time cutting fossil-fuel use and emissions behind climate change. Simples!

Better energy-efficiency

Taking this a step further, AI-powered algorithms can process huge data sets from sensors and devices embedded in buildings, industries, and transport. As such, they can track energy demand trends, pinpoint inefficiencies, and suggest strategies to reduce energy waste.

Renewable energy and smart grid management

Net-zero depends on a successful transition from fossil-fuels to renewable energy. Advanced AI algorithms working with smart-grids and energy storage systems that may even be able to take weather patterns into account can balance the overall output and distribution of renewable energy to stabilise demand with supply.

Letting the sunshine in

As an example, when clouds move over solar panels, power generation can fall off quickly. Open Climate Fix (https://openclimatefix.org/) working with the National Grid (https://www.nationalgrid.com/) uses AI to provide more accurate solar forecasts.

This is where big data counts. Using readings from more than 25,000 UK-based solar power systems, it predicts how much power will be produced, and says short-term forecasts can cut carbon emissions by some 100,000 tonnes annually.

Material world

One other key area where AI has environmental gains is the development of new sustainable materials. Here, it can design materials with specific properties like greater strength or reduced weight that benefit home-building, construction, and even aerospace.

Made from renewable sources, new materials can both reduce fossil fuel use and minimise the environmental impacts of their own manufacture, particularly when recycled.

Monitoring and predicting

As mentioned above, one of the big bonuses we can expect from AI is a greater ability to monitor, and then respond to extreme and erratic weather patterns of the type that have caused extensive flooding across the UK in recent years. Air and water quality can also be monitored in real-time.

Farming, sustainable food, and ‘tree guardians’

Precision agriculture is another application where AI is helping farmers to cut their fertiliser and pesticide use – resulting in healthier crops and lower environmental contamination. AI can also optimise food transport routes and fuel consumption, leading to lower emissions and better air quality.

Sensors, drones, and AI algorithms on farms also monitor soil conditions, water usage, crop health and resource use, cutting waste and minimising negative footprints. It has been estimated this could cut greenhouse gas emissions in agriculture, water, energy and transport by 4% by 2030.

And if trees are in danger of falling in the forest …

Yes … they can now send out warning based on sounds heard around them! Trees are major carbon dioxide sinks. When they are felled carbon dioxide escapes into the atmosphere. To curb illegal land clearing, one inventive AI solution has been to attach acoustic monitoring sensors to trees that ‘eavesdrop’ on surrounding forest sounds and send audio signals to the cloud.

A machine learning model has been trained to recognise sounds linked to illegal logging, such as a chainsaw or engine noises. Warning are then sent to rangers. Some 600 ‘tree guardians’ have been installed in 35 countries – from Brazil to Indonesia, Congo and the Philippines – where they have collected more than 107 million minutes of audio data from more than 400,000 hectares of land.

More information from Enzygo

Clearly, the use of AI with data is both fascinating and very promising. If you would like to discuss any of the issues above in more detail, please feel free to contact me directly.

Scott Dawson, Principal Consultant, Enzygo Ltd