Read more
HSO in Public Sector
More accurate and reliable traffic information for Hamilton City Council
Traffic Machine Learning proof of concept showed that a camera-based solution is low cost, high accuracy, real-time and permanent, and can provide significantly more information.
Hamilton is growing fast and Hamilton City Council is focussed on making sure the city’s infrastructure is aligned to support that growth, whilst ensuring Hamilton is a liveable city for members of the community. Hamilton City Council (HCC) manage almost all roads across Hamilton, handling a vast and growing number of vehicles per day. Current traffic management industry practices include high-cost, low-accuracy solutions involving pressure sensors. These solutions are slow to install, rely on devices that must be moved around the city, and encode very little information.
Smart Hamilton programme
Through the Smart Hamilton programme, Hamilton City Council and Aware Group (an HSO company) have been working together to trial a number of Artificial Intelligence (AI) technologies to enhance their capability to manage the city for today, and into the future. One of these initiatives is the Traffic Machine Learning proof of concept, which involves trialling automated classification and counting of vehicles in real-time.
In contrast with current approaches, the Traffic Machine Learning proof of concept showed that a camera-based solution is low cost, high accuracy, real-time and permanent, and can provide significantly more information including specific types of vehicles and driver behaviour.
Customer challenges
In their 2018-2020 10-Year Plan, Hamilton City Council has committed to developing “a transport programme to improve safety, congestion and transport choice (walking, cycling and public transport)” as well as “taking up opportunities afforded by innovation, insight and technology to realise savings of $94M.”
To improve transport safety and congestion, Hamilton City Council requires information on vehicle behaviour in order to effectively make core decisions around city infrastructure such as road building and maintenance, road speed limits, city zoning, and many other issues.
Current industry practices to obtain traffic information rely on technology that costs upwards of $70,000 per year, while generally only providing four week-long snapshots of up to 150 locations per year. These ~600 data points are then extrapolated in order to make critical decisions that affect the entire city. This Smart Hamilton initiative is tasked with improving the traffic data available and thereby the decisions made.
“In addition, the use of computer vision opens up additional avenues of use, including counting foot and cycle traffic, as well as recording driver behaviour such as intersection use and queuing.”
Identified requirements
As innovation specialists with vast experience in Computer Vision and Business Intelligence, Aware Group is perfectly positioned to support Hamilton City Council to achieve their goals.
Aware Group worked with Hamilton City Council to identify the requirements of their transport goals then proved that a machine learning model that uses computer vision to analyse existing camera footage in real-time will be able to significantly improve traffic data across Hamilton.
This model produces a dataset including the number, type, and direction of every vehicle visible to the camera resulting in more accurate and higher fidelity information than the previous pressure-based sensors could provide. In addition, the use of computer vision opens up additional avenues of use, including counting foot and cycle traffic, as well as recording driver behaviour such as intersection use and queuing.
Customer Outcomes
Through Traffic Machine Learning, Hamilton City Council will have improved traffic data through an increased dataset with real-time information on the number of road users. This data allows them to make more accurate and informed decisions, thereby improving transport safety and congestion across Hamilton and improving liveability.
As a result of this Proof of Concept, Hamilton City Council is currently in the process of organising a larger Smart Hamilton computer vision programme of work. This includes the basic traffic counting and classification as well as identifying other use cases where the cameras involved can add value by running additional Machine Learning models.
In addition, Hamilton City Council has initiated a process of identifying how computer vision can be used by in an ethical, legal, and efficient manner to improve the city while maintaining citizens’ privacy – an initiative Aware Group is proud to support.
Aware Group is excited to work with Hamilton City Council on future initiatives under the Smart Hamilton programme to show how innovation and technology can be embraced to serve the local community.