Automation

Samira Namin, Aurecon - Smart Pedestrian Crossing

Samira Namin

Lead ITS Engineer | Aurecon

Samira is a Lead ITS Engineer in Integrated Transport and Mobility team of Aurecon.  She comes from a security/surveillance background and has been working in transport industry since 2003. Her working experiences in ITS cover Brisbane City Council, Department of Transport and Main Roads and WSP.

Smart Pedestrian Crossing

This abstract focuses on pedestrian automated detection to achieve future mobility and innovations around signalised intersections and is prepared based on recent Safer Roads Sooner (SRS) program owned by Department of Transport and Main Roads (TMR).

Single-stage pedestrian crossings provision on all intersection legs is a default recommendation at urban areas. It intends to minimise pedestrian delay times as far as possible considering intersection geometry. Queensland research found pedestrian compliance is highest where delay time is between 60-90 seconds.

There is an almost 50% decrease in compliance for the delay times exceeding this ratio. Signals incorporating pedestrians detection technology provides reduced delay to motorists, reduced cycle time and improved level of service (LOS).

Dual optical system in the detector is designed to detect and monitor waiting pedestrians/cyclists across the road and ensures the crossing phase is only called when the pedestrians/cyclists are present.

The Automation and ITS is a versatile topic aiming to balance various transportation rigidity. On the one hand, an easier, safer, integrated and cheaper travel enabled by modern technology is recommended and on the other hand, there is a risk of more vehicles, congestion and impacts on space.

The way goods, vehicles and people move from A to B is evolving; and it is not only about autonomous and connected vehicles. The entire mobility ecosystem including engineering rise to climate challenge, along with our expectations, is changing the world for better.

Developments such as technological infrastructure, ridesharing, autonomous vehicles, micro-transit shuttles, E-scooters, truck platooning, and drone delivery benefit individual travellers, governments, and businesses.

Craig Wooldridge, Main Roads - Developing Automated Intelligence to Access Conflicts and Behaviours on Shared Paths 

Craig Wooldridge

Manager Project Development - Traffic & Safety| Main Roads

Craig Wooldridge has extensive experience in the transport planning and traffic engineering professions gained through 39 years in the industry. 

Craig is a Past Chair of Cycling and Walking Australia and New Zealand, a Past National President AITPM and is currently a WestCycle Board member.

Co-author(s):

Chao Sun | Planning & Transport Research Centre (PATREC)

Chao Sun is a Senior Research Fellow at the Planning and Transport Research Centre (PATREC), a collaboration established in 2003 between three WA universities (UWA, Curtin, and ECU) and WA's transport portfolio including Main Roads.

Before joining PATREC, Chao worked as a Senior Consultant at AECOM. His unique combination of a broad range of industry and academic experience gives him an advantage in bridging the gap between the two. 

Since 2016, he has been actively involved in research projects worth over AU$4.5 million, out of which he has led (either as the project leader or technical lead) projects of over AU$2.1 million, most of which are industry-funded. By applying the agile development process to mitigate risks in R&D, he and his team successfully delivered tangible research outcomes with both academic and industry value and a high level of client satisfaction.

He focuses on applying data science and machine learning techniques to modernise the transport industry. One of their AI-based solutions is being commercialised by UWA. 

James Pearse | Hatch

James Pearse is an experienced transportation consultant who has worked in the public, private and not-for-profit sectors. James holds a Bachelor of Civil of Engineering from UWA and a Master of Urban and Regional Planning from Curtin University.

Key areas of interest for James include active transport, public transport and road safety. James leads Hatch RobertsDay's Perth-based transport advisory team, working closely with the firm's land use planners, urban designers, economists and placemakers.

Developing Automated Intelligence to Access Conflicts and Behaviours on Shared Paths 

Perth's strategic shared path network is made of high-standard paths shared between pedestrians and cyclists and more recently e-rideables. It the backbone of the active transport network and a number paths carry over 1,000 cyclists per day and high numbers of pedestrians.

Although they are generally safer than roads, crashes and near misses still happen with risk levels increasing due to the increasing numbers of e-rideables. The most common concerns include cyclists/e-rideables travelling at a high speed and not giving way to pedestrians, especially around train stations.

Different types of interventions have been applied but there is a lack of evidence on their effectiveness because collecting pedestrian and cyclist data has been technically challenging. 

Our project addressed these problems by applying Video Analytics (VA) to automate the analysis of video surveys for pedestrians, cyclists/escooter/skateboards, and cars. VA involves the automatic extraction of data from videos using modern machine learning techniques, especially deep learning.

It has only recently become practical because of rapid advances in algorithms and computer hardware. Its non-invasive nature and the richness of extracted information make it an attractive technology for pedestrian, cyclist and e-rideable surveys. As a pilot, 29 videos from nine sites were collected. 

The software was employed to automatically identify and track objects of interest such as pedestrians, cyclists, e-rideables, skateboards, and vehicles across each site. The extracted data provided various insights surrounding path user movements, volumes, speeds, give-way compliance, and conflicts.

It provides a visual representation of their movement patterns and highlights areas with the most opposing movements, which are denoted by a bright white light. The software also automatically generates summaries of moments with potential safety concerns and heat maps displaying movements. 

David Barnes, TTM Consulting Pty Ltd - Flip the script on Origin-Destination Surveys

David Barnes

Director | TTM Consulting Pty Ltd

David commenced with TTM in 1997 and has over 20 years experience in transport data studies. He has been a Board Director and part-owner of TTM since 2010. 

Flip the script on Origin-Destination Surveys

Restricted number plate surveys as a key input to Major Project decisions:

Number plate surveys of various types have traditionally been restricted by expensive and lengthy analysis processes. Even with the adoption of video-based surveys, the process of number plate extraction or processing from videos has been completed by either manual processes (i.e. data centre) or lengthy post-survey optical character recognition software. 

The high cost related to this type of survey has led to project data being driven by a function of cost, rather than by the dataset that would best meet the study objectives. Origin-Destination survey data (using number plates) is one of the best inputs to major projects and decisions, therefore it is not ideal for this input to be driven primarily by cost. 

As an example, data may be requested for one day during peak hours. Yet, it is unknown whether this short term data is representative of typical traffic conditions. 

Real-time technology in short term surveys:

The latest technology enables Smart and Connected ANPR cameras to process number plates on board, upload data to a server, and allow for matching of plates within a near-real-time window. 

The new technology eliminates the cost-intensive, lengthy, post survey processing and opens up the possibility for cost-efficient, long-term Origin Destination Surveys, delivered within day/s of the field survey. 

Benefits:
The primary benefit of this latest technology is to increase the quality of data available for input into major project decisions by removing the cost constraints on the acquisition of number plate data.