4 minute read
ADVANCES IN AI ARE BUILDING SMART CITIES
By Dr. Adam Mowlam, Manager Smart Cities,
Artificial intelligence
(AI)
is accelerating
City of Greater Geelong
Quickly With The Potential For It To Change
the fabric of many sectors – and possibly the world as we know it. One sector in particular that can deliver more streamlined services thanks to AI is local government, with the possibility for improved productivity encouraging many councils to adopt the technology.
Introduction To Ai In Local Government
The potential benefits of AI and Robotic Process Automation (RPA) adoption in the Australian local government sector includes improved efficiency and productivity, cost savings, and the ability to provide better, more personalised services to citizens.
AI can help automate routine administrative tasks, such as validating information required for permits and registrations, and enable governments to make more evidenced-based decisions by analysing large amounts of data at scale.
This article considers the use of generative AI-based technologies, such as ChatGPT, computer vision, and natural language processing, and its potential to transform the way government services are delivered, by making them more accessible, convenient, and efficient for citizens.
The Local Government Act 2020 And Machine Readable Local Laws
Local governments have the authority to create laws and regulations to address local issues. Here, Council's local law on camping is utilised as an example to demonstrate the potential influence of AI on local government.
Under the local laws prescribed by each council, camping may be allowed under certain conditions, including the requirement to obtain a camping permit. The regulations governing camping differ between councils and take into account various factors, such as land ownership (e.g. council or private), availability of facilities (such as washing, toilets, and laundry), and the potential impact on the local community.
Ongoing permission to camp may depend on factors such as campsite cleanliness, environmental impact, adherence to campfire regulations, and payment of fees.
The interpretation of local laws differ between councils, as can the breadth of subject matter; for example, rural councils may have laws regarding shipping containers and livestock, while metropolitan councils may have laws for intruder alarms and shopping trolley control. Many councils also publish a procedural manual that guide the steps and responsibilities involved in implementing local laws, serving as a reference for those enforcing and complying with them.
Rules-as-Code (RAC) is a method of transforming laws and regulations into a computer-readable format, with the aim being to structure content logically for the purpose of enabling automation.
The more structured and consistent the local law, the more effectively AI can use them to generate accurate, relevant answers. The existence of ambiguity in language will generate uncertain or conflicting responses requiring subjective analysis by a human. Figure 1. illustrates a number of possibilities, conditions and requirements for digital-ready regulations.
Ruleset and definition needs to be defined (larger block, greater allowance)
Data available from building permits or on-site residential accommodation
Ruleset needs to be defined - and relevant data accessed (land zone, proximity to neighbours, land slope)
Ruleset needs to be defined
This would prevent automation
More data will need to be made open for external AIpowered tools to access – for example, a test of council land ownership required to determine whether a permit can be granted for overnight camping.
Benefits And Risks Of Implementing Ai
To provide better customer experiences with AI, councils need to adopt digital-age strategies for designing web content that enables seamless sharing of information between government entities (G2G) and government and businesses (G2B).
This involves standard data formats, common vocabulary, open APIs, and increased use of metadata. A RAC approach to local laws and government services will support AI and offer benefits, given the nature of service delivery across government tiers.
This will provide four key benefits:
∞ Optimising administrational overheads whilst expediting citizen access to services and increasing resident connectivity to community in a cost effective manner
∞ The machine readability of local laws improves councils' ability to deliver clear, understandable, responsive and adaptable rules for better administration
∞ Making government rules machine-readable for the public can enable the development of useful services by other agencies and businesses
∞ Alignment with priorities and initiatives of other government tiers, such as the digital economy and increasing transparency and accountability in decision-making
These new approaches, driven by the possibilities of AI, are not limited to text and language, they can also be used for spatial analytics, for example, digital twins which are digital representations of the built and natural environment can be utilised for simulation and analysis purposes.
In the context of building approvals, digital twins can be used to conduct automated three-dimensional building compliance checks against building codes, such as setbacks and car parking requirements.
By combining digital twins and machine readable code, architects, engineers, and building designers can ensure that the building designs are compliant with local building codes and regulations, while also optimising other key performance indicators such as the building's energy efficiency. The same technology stack can be applied to non-text local laws, such as regulations regarding tree branches overhanging footpaths.
As with any emerging technology, AI carries its own set of risks and concerns. These include potential biases embedded in AI algorithms, opaque decision-making processes, cyber threats, and privacy breaches. Additionally, the limited number of technology providers offering infrastructure and services can lead to elevated costs and systemic risks. Thus, it is imperative for governments to take action and implement measures to mitigate these potential adverse effects.
Strategies For Adopting Next Generation Technology
The advancement of AI presents a significant opportunity for councils, but ensuring local legislation for machines cannot be delivered without collaboration and leadership.
As digital leaders, we have a chance to alter what the future of local government may look like, and it begins with boosting awareness and promoting innovation as a sector-wide objective.
In terms of key steps, the following is recommended:
∞ Cooperation and knowledge-sharing at the national, regional, and international levels, requiring the coordination of legal and regulatory actions
∞ Creating a central repository of machine-readable local laws that can be easily accessed and shared between councils
∞ Facilitating collaboration and sharing of resources and knowledge between councils and AI experts
∞ Supporting research and development in the field of AI and machine-readable laws to ensure that councils have access to the latest technologies and techniques
∞ Establishing a working group that brings together experts in AI, and council leaders, to develop best practices and guidelines
∞ Providing funding for councils to hire experts in AI and machine-readable law to help them develop these laws
∞ Offering training and education programs for council officers on the use of AI and machine-readable laws
In Conclusion
The integration of generative AI technology holds great promise for the sector given the analog nature of council operations, and these advancements have the potential to revolutionise service delivery and improve overall customer experience.
Clear, understandable, and machine-readable legislation should be prioritised to facilitate improved automation, effective administration and accessibility.