adaptive traffic control system - efftronics
TRANSCRIPT
10101010101010101
10101010101010101
Adaptive Traffic Control SystemB r i n g i n g I n t e l l i g e n c e t o I n t e r s e c t i o n s
A traffic phase is defined as the green, change, and clearance
intervals in a cycle assigned to specified movement(s) of traffic.
PHASE
A series of non-conflicting phases which run together.
STAGE
A cycle is defined as the total time to complete one sequence of
signalization for all movements at an intersection.
CYCLE
The difference between the GREEN initiation times at two adjacent
intersection.
OFFSET
Actual time available for the vehicles to cross the intersection.
EFFECTIVE GREEN TIME
TRAFFIC DOMAIN
The DoS of an intersection is a measure of how much demand it is
experiencing compared to its total capacity.
DEGREE OF SATURATION (DoS)
TRAFFIC DOMAIN
Signal timings derived from the statistical data
Duration and order of all green phases are fixed
Cannot respond to real-time demand
Requires frequent re-configurations and updates
Traditional Signal System
Low level of lane discipline
High mix of traffic
Lack of local expertise
Poor junction geometry
Lack of expertise
Power and Network connectivity interruptions
Challenges in INDIA
Solution
A solution for a smooth and safer journeys
Adaptive Traffic Control System
What is ATCS ?
Adaptive Traffic Control System adapts to real time
traffic patterns to optimize the traffic flow by
dynamically change the green split timings.
Detectors
monitors
traffic
Algorithms
compares to
current plan
Change
timings if
needed
1
2
3
ATCS
An ATCS uses Advanced Machine Learning algorithms that adjust
❖ Cycle length
❖ Phase sequence to
Minimize delays Reduce the number of stops Decrease the travel time
OBJECTIVES
Reduce Accident Rates Increase Travel Speeds Reduce Stops, Delays & Queues
Increase Operational
Efficiency
Real Time Information management
Create a platform for sharing traffic to other systems
Where is
ATCS most effective?
Where frequent and unpredictable
changes of demand, events, weather
situations, etc., creates unexpected
fluctuations in the system
Where unpredictable traffic changes
results in delays or stops that cannot
be addressed by conventional signal
timing.
ATCS does not necessarily
solve the capacity problems
of over saturated corridors.
Adaptive Traffic Control System
Completely Designed for Indian Traffic Conditions
Bringing Intelligence to Intersections
Employs advanced Machine Learning Algorithms
Traffic data from junctions is consolidated in a central traffic system
Dynamically adapts to changing traffic conditions
Junction green-green synchronization
1. Vehicle detector
6. Web Interface
7. Adaptive Algorithm
8. Real-time Reports
9. ML based forecasting
10. API Services
4. Master Controller
5. Edge Application
2. LED Signal Lamps
3. Countdown Timer
ATCS Controller
Slave ControllersSerial Wired / WirelessControl Logic & Split
Optimization
Signal Aspects
Switching Logic
Vehicle Detectors
Communication Interface
Ethernet / GPRSVehicle data, Stage timing, Saturation etc.
Configured Data[Routes, Scheduled Timings, One-ways, Preferences etc.]
KPIs calculation for intersection, approach &
corridor
Configured Data[min/max phase length, Cycle lengths, Pedestrian timings]
Current Signal Timings
Route/Corridor Selection
Critical Cycle Time & New Phase Time Calculation
Offset Calculation
Cycle & Split Calculation
New Timings Calculation
ATCS Algorithm
Vehicle Detection
Phase Sequence Information
Way Volume Calculation
Saturation Flowrate Calculation
Online Split Optimization Learning Data
Traffic Intersection
Machine Learning AlgorithmStage Optimization / Performance Index
Vehicle data, Stage timing, Saturation etc.
Central Server
Timings TranslationPreferences & Observations
Offset OptimizationGroup Cycle Length
CalculationKPIs calculationWeights & Bias
New timings & Preferences
Machine Learning AlgorithmTraffic Forecasting
Vehicle Data, Events, KPIs, Forecasting Data
Integration APIs
Web interface
Switching Logic
Mode of Operation Stage timings execution Vehicle Detectors Input Pre-emption Cycle Time
Fixed Time (Pre-timed) As per site specific time-table. Ignored Not made on any stage Constant
Vehicle Actuation with Allstages Pre-emption
As per demand from vehicle detectors YesFor all demand actuatedstages
Variable
Semi-ActuationAs per demand from vehicle detectors forvehicle actuated stages. Remaining stagesexecute maximum green time configured.
YesFor all demand actuatedstages
Variable
Vehicle Actuation with Fixed Cycle Length
As per demand from vehicle detectors YesFor all demand actuatedstages except for prioritystage
Constant
Adaptive Traffic ControlAs per demand from vehicle detectors forthe stage and cycle time computed bycentral algorithm
YesFor all demand actuatedstages except for prioritystage
Variable
Manual (Power Down – Blank,Amber Blinker, Green corridorand Police Hurry Call)
As per setting, only one defined stage Ignored Not made on any stage Constant
Transit Signal Priority forPassenger Buses / Emergencyvehicles
As per setting, only one defined stageIgnored Not made on any stage Constant
Fixed Mode (For 2 cycles)
Vehicle Actuated
Mode
Semi Actuated
Mode
ATCSMode
Powered UP
cameras.connected = 0
cameras.connected > 0 .and.
< cameras.config.countcameras.connected = 0
cameras.connected = cameras.config.countcameras.connected > 0
.and.< cameras.config.count
central.connect = true
central.connect = false.and.
(cameras.connected > 0.and.
< cameras.config.count)
central.connect = false.and.
cameras.connected = cameras.config.count
cameras.connected = 0
central.connect = true.and.
cameras.connected = cameras.config.count
Manual/Priority
Mode
manual.control = true.or.
vehicle.priority = true
manual.control = false.and.
vehicle.priority = false.or.
timeout.event = true
central.connect = true&
cameras.connected = cameras.config.count
ATCS ELEMENTS
API
Services
Adaptive
Algorithm
ML based
Forecasting
Edge
ApplicationMaster
Controller
Countdown
timers
LED signal
lamps
Vehicle
Detectors
Web
Interface
Real-time
reports
1
6
8
4
7
10
3
5
2
9
ATCS
MASTER CONTROLLER
Inputs derived from Cameras/Sensors/Radar will allow the ATCS Controller to
adaptively adjust the junction timing pattern and communicate the same to the
next junction and the control room.
VEHICLE DETECTORS
Detects traffic flow, headway, average speed, occupancy and queue
length by using in-built advanced machine learning algorithms.
Cover up-to 1-3 lanes at a time and has 90% detection accuracy
Any adaptive traffic control system relies upon good detection of the current conditions in real-
time in order to allow a quick and effective response to any changes in the current traffic situation.
LED SIGNAL LAMPS
EN 12368 Compliant
Premium quality LEDs
10+ years working life
Designed as per Indian Conditions
COUNTDOWN TIMERS
EN 12966 Compliant
Signal time indications to road commuters
10+ years working life
Designed as per Indian Conditions
ADAPTIVE ALGORITHM
The well advanced ATCS algorithm fulfil the ATCS main objectives
based on Indian traffic conditions. ATCS algorithm determines
optimized red-green phases of traffic signals in order to achieve
junctions green-green synchronization.
ATCS
EDGE APPLICATION
Offers a high-level framework which provides a single access
point for all the component systems and support for the whole
life cycle of a system: implementation, operation, updating and
planning.
Process Vehicle detector data
Capture and store all signalling events
Able to run junction in VA/Fixed modes when networks fails
Auto data backup & recovery mechanism
In built edge computing capabilities
ATCS
WEB INTERFACE
Intelligent Web browser-based access, requiring no local setup on
the control centre with hierarchy based secured login to operators
and city managers for their dynamic and strategic planning in real
time.
Core traffic management application in ATCS
Secure user level based authentication
Centrally monitor/control/configure junctions from one place
Interactive UI to use any novice user without prior domain knowledge
Inbuilt monitor, control, reports modules
ATCS
WEB INTERFACE
Live monitoring for quick insights & decisions
Junction mimic screen
Junction Configuration
Junction Configuration
Signal Configuration
Day plans Configuration
Green Corridor Feature
REAL-TIME & HISTORICAL
ADAPTIVE REPORTS
Exceptionally flexible and user-friendly graphical interface which
is also multilingual. The clear graphics allow rapid and intuitive
interpretation of the real-time status of the network.
KPI Reports – Time space Diagram
MACHINE LEARNING BASED
TRAFFIC FORECASTING
ATCS uses machine learning algorithms to analyse real-time traffic
data to determine signal timings that are optimal for existing
traffic conditions along the corridors.
MACHINE LEARNING BASED
TRAFFIC FORECASTING
Predictive analytics for prescriptive maintenance
API SERVICES
This gives all applications access to high quality data, which is
available to operators and city managers for their strategic
planning. The DBs themselves remain independent in order to
keep the system open and flexible.
High quality data set
Get quick junctions signal & health events in JSON with simple API call
Various data formats for heterogenous utilisation like RLVD, VMS info etc.,
Get live traffic situations to display at Information sign boards
Quick alerts with faster response
BENEFITS
Bus priority compensation and emergency vehicle pre-emption.
Integration to share the predicted forecasts to different dissemination
systems.
Suggestions to future infra requirements based on the real time traffic
density.
Power saving by regulating the intensity based on ambient light
sensor.
Standard Models Comparison
SCOOT SCATS efftronics ATCS
Upstream detection Downstream detection Downstream detection
Centralized system Distributed system Distributed system
Fixed traffic region Adjustable region Adjustable region
Fallback - fixed Fallback - VA Fallback - VA
Model based [Time generated] Algorithmic [Plan selection] Algorithmic [Time generated]
User Interface was little bit complexlacks user-friendly interface
features to support day-to-day operation
user-friendly interface features to support day-to-day operation
Doesn’t make real-time adjustments as traffic volume
changes
Does not have the predictive capability
Has predictive capability
Limited scalability Limited scalability High scalability
Not fit for developing nations Not fit for developing nations Designed for developing nations
SCOOT - Split Cycle Offset Optimisation TechniqueSCATS - Sydney Coordinated Adaptive Traffic System
CASE STUDY
Gandhinagar Smart CityGujarat, INDIA
Gandhinagar Smart City, India’s first smart city, and the second planned
city after Chandigarh. Efftronics ATCS played a major role in solving the
most vulnerable traffic problems and leads to the Vision of the city of
being an equitable urban centre that provides a high quality of life to all its
citizens.
Reduce travel times by
12-20%
20-40%Less waiting time
10-20%more friendly to
environment
Reduce Number of stops
in a corridor by
15-30%
Thank You
@efftronics_ltd /efftronics /company/efftronics
Follow us
www.efftronics.com