Prepared for the Architect of Tech — Transport for NSW (TfNSW)

Asset Software & Tech Matrix: Roads, Trains & Ferries
Asset Software & Tech Matrix: Roads, Trains & Ferries

1. Next-Generation Traffic Management & Signal Orchestration

1.1 SCATS Evolution: From Inductive Loops to SCATS 7 & Deep Reinforcement Learning (DRL)

  • Sovereign Heritage: Invented in NSW (DMR/RTA, now TfNSW) in the 1970s, SCATS (Sydney Coordinated Adaptive Traffic System) controls >65,000 intersections across 180+ cities in 28+ countries (<a href="https://www.scats.nsw.gov.au/” target=”_blank” rel=”noopener noreferrer” class=”inline-ext-link”>https://www.scats.nsw.gov.au/` ↗).
  • Classic SCATS Mechanics: Sub-surface inductive loop detectors at the stop-line measure Degree of Saturation ($DS$)—the ratio of effectively used green time to total available green time—alongside reconstituted traffic volume to dynamically adjust Cycle Length ($C$), Phase Splits ($g/C$), and Subsystem Offsets between adjacent intersections.
  • SCATS 7 & Multi-Agent Deep Reinforcement Learning (MARL):
  • Next-gen SCATS decouples physical roadside traffic signal controllers from cloud and edge AI optimization engines via open, low-latency APIs.
  • Through iMOVE CRC and UNSW/USyd research pilots, Deep Reinforcement Learning (DRL) agents ingest continuous spatial-temporal trajectories from Connected Vehicles (CAV probes), radar, and computer-vision cameras—seeing approaching platoons 300–500 meters upstream rather than only at the stop-line loop.
  • Multi-Objective Reward Function:

$$R_t = -left( w_1 sum ext{QueueLength} + w_2 sum ext{VehicleDelay} + w_3 sum ext{BusDelay}_{ ext{priority}} + w_4 sum ext{PedestrianWait} + w_5 sum ext{CO}_2

ight)$$

  • Verified Metrics: Classic SCATS reduces travel times by 15–28%, stops by 20–40%, and fuel use by 12–15%; fusing even low-penetration (5–15%) CAV trajectory data with DRL signal control delivers an additional 10–20% reduction in intersection delay and 30%+ queue length reduction during non-recurring incidents.
Hangzhou & Global AI Traffic Orchestration
Hangzhou & Global AI Traffic Orchestration

1.2 Edge-Compute Computer Vision & Surrogate Safety Analytics

  • Roadside Edge Architecture: Ruggedized industrial AI nodes (NVIDIA Jetson Orin / Hailo-8) inside traffic signal cabinets process CCTV, uncooled thermal LWIR, and 4D millimeter-wave (mmWave) radar streams locally (sub-100ms inference, zero cloud video backhaul cost, GDPR/privacy-compliant edge anonymization such as VivaCity sensors deployed across NSW).
  • Capabilities:
  • Proactive Near-Miss Detection: Instead of waiting 5 years for fatal crashes to accumulate at an intersection to identify a blackspot, edge vision calculates continuous Surrogate Safety Measures (SSMs)—specifically Time-to-Collision ($ ext{TTC} < 1.5 ext{ s}$) and Post-Encroachment Time ($ ext{PET} < 1.0 ext{ s}$) between turning vehicles, cyclists, and pedestrians.
  • Dynamic VRU & Freight Phase Extension: Automatically extends pedestrian walk phases when an elderly person or wheelchair user (8-to-80 Cities design) is still in the crosswalk, or holds a green phase for an approaching 68-tonne B-double freight truck to prevent dangerous hard-braking or red-light running.
  • Tunnel Incident Detection: Instantaneous detection of stopped vehicles, wrong-way drivers, smoke, or spilled cargo inside WestConnex, NorthConnex, Sydney Harbour Tunnel, and Western Harbour Tunnel.

1.3 C-ITS / V2X (5G NR-V2X PC5 Sidelink) & Smart Motorway Algorithms

  • Protocol Evolution: Moving from IEEE 802.11p / DSRC in the $5.9 ext{ GHz}$ ITS band toward 3GPP Release 16+ 5G NR-V2X:
  • PC5 Direct Sidelink: Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) direct radio link with $<5 ext{ ms}$ latency that operates even without cellular tower coverage.
  • Uu Cellular Interface: Long-range Vehicle-to-Network (V2N) cloud broadcasts for regional hazard warnings.
  • Standard Messages: ETSI ITS-G5 CAM (Cooperative Awareness Message), DENM (Decentralized Environmental Notification Message), SPaT (Signal Phase and Timing — telling the car's dashboard or ADAS the exact countdown to green), and MAP (digital intersection topology).
  • NSW Deployment: TfNSW's CITI (Cooperative Intelligent Transport Initiative) in the Illawarra equipped heavy trucks and traffic signals with Cohda Wireless V2X units (<a href="https://www.cohdawireless.com/” target=”_blank” rel=”noopener noreferrer” class=”inline-ext-link”>https://www.cohdawireless.com/` ↗), backed by Austroads' National C-ITS Public Key Infrastructure (PKI).
  • Smart Motorway Shockwave Damping (M1, M4, M7):
  • Combines Coordinated Ramp Metering (HERO algorithm) with Dynamic Variable Speed Limits (VSL) to keep mainline density below the critical breakdown threshold ($sim 28 ext{ veh/km/lane}$ or $sim 2,000 ext{ veh/hr/lane}$), preventing "phantom traffic jams" (stop-and-go kinematic shockwaves).

2. Predictive Road Asset Maintenance, Robotics & Smart Pavements

Highway 15-Second Solar Drone-in-a-Box
Highway 15-Second Solar Drone-in-a-Box

2.1 Crowdsourced Connected Vehicle Telemetry (Compass IoT) & Mobile LiDAR

  • Compass IoT (Pavepoint) — Sydney Sovereign Innovation:
  • Rather than relying solely on specialized laser survey trucks (NTRO iPAVE) that inspect regional roads once every 1–2 years, Compass IoT (<a href="https://www.compassiot.com.au/” target=”_blank” rel=”noopener noreferrer” class=”inline-ext-link”>https://www.compassiot.com.au/` ↗) ingests high-frequency (1–10 Hz) vertical acceleration ($a_z$), pitch/roll rates, ABS/ESC activations, and GPS trajectories from hundreds of thousands of everyday connected cars and commercial fleets.
  • Normalizing across vehicle suspension classes estimates the International Roughness Index (IRI, in m/km) continuously—detecting pothole formation and flood damage within hours of heavy rain across the entire NSW road network.
  • High-Speed Mobile LiDAR & Vision Transformers:
  • Survey vehicles equipped with dual laser scanners (Riegl VMX / Trimble MX9, $>2 ext{ million points/sec}$) and downward-facing 4K line-scan cameras run Mask R-CNN / Vision Transformer models to classify sub-millimeter alligator cracking, rutting, ravelling, and guardrail/sign retroreflectivity loss automatically into Digital Twin NSW.

2.2 Self-Healing Asphalt & Dynamic Inductive EV Charging Lanes (Electreon)

  • Self-Healing Pavements:
  • Micro-encapsulated Bio-Rejuvenators: Polymeric microcapsules mixed into bitumen rupture when micro-cracks form, releasing maltene oils that seal the crack via capillary action.
  • Induction-Heated Conductive Asphalt (TU Delft): Embedding conductive steel wool fibers into porous asphalt allows a maintenance vehicle carrying an induction coil to heat the binder locally to $85–110^circ ext{C}$, melting bitumen to close micro-cracks and extending road surface life by 30–50%.
  • Dynamic Wireless Power Transfer (DWPT — Electreon):
  • Copper induction coils buried $8–10 ext{ cm}$ beneath asphalt (operating at $85 ext{ kHz}$ magnetic resonance per SAE J2954) energize only when an authenticated EV bus or heavy truck passes overhead (<a href="https://www.electreon.com/” target=”_blank” rel=”noopener noreferrer” class=”inline-ext-link”>https://www.electreon.com/` ↗).
  • Deployed in Sweden, Israel, USA (Detroit), Germany, and France's A10 motorway ($200–300 ext{ kW}$ dynamic transfer), allowing heavy electric trucks and buses to downsize onboard batteries by 50–70%.
Inductive Charging & Smart Pavement Architecture
Inductive Charging & Smart Pavement Architecture

2.3 Autonomous Roadworks Robotics & Bridge Structural Health Monitoring (SHM)

  • Autonomous Cone Laying Machines (ACLM) & Crack Robots:
  • Automated Cone Trucks (BladeRunner / Kier / Highways England and Australian trials) deploy and retrieve traffic cones at $15–20 ext{ km/h}$ via robotic arms, keeping human workers out of live $110 ext{ km/h}$ lanes.
  • Robotiz3d (ARU) uses stereo vision and 3D LiDAR to autonomously detect, clean, and fill road cracks with bitumen sealant.
  • Structural Health Monitoring (SHM) for Iconic Bridges & Tunnels:
  • Fiber Bragg Gratings (FBG) & Distributed Strain/Acoustic Sensing (DSS/DAS): Optical fibers bonded to bridge girders and suspension cables measure wavelength shifts ($Delta lambda_B$) to monitor micro-strain ($mu

arepsilon$) and temperature immune to electromagnetic interference.

  • Acoustic Emission (AE) Wire-Break Sensors: Piezoelectric sensors ($20 ext{ kHz}–1 ext{ MHz}$) detect the exact acoustic elastic wave of a single high-tensile steel wire snapping inside a bridge cable or post-tensioned concrete box girder years before visible structural sagging occurs.
  • UTS Robotics on Sydney Harbour Bridge: Autonomous climbing grit-blasting robots (CROC) inspect and strip lead paint inside confined bridge arches safely.