ADAPTIVE TRAFFIC LIGHT CYCLE TIME CONTROLLER USING .

ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume IV-4/W7, 20183rd International Conference on Smart Data and Smart Cities, 4–5 October 2018, Delft, The NetherlandsAlthough road traffic authorities in India try to mitigate theproblem by switching different fixed time traffic signal timingon basis of the time-of-day plan, this is ineffective as traffic isgenerally random for the larger time period, which is taken to beon an hourly basis. Traffic congestion being a stochastic processcan be mapped more accurately for the small time- period,therefore, managing traffic light will be better if time- periodconsidered for congestion checking is small. Also, the trafficcongestion build-up curve states that build-up of congestion is ashort extent temporal and spatial process but once built-up ittakes hours to disperse. This leads to an irregular traffic delayduring transit in the urban areas (Sen, Rijurekha, et al., 2013).Due to this, not only the vehicles get trapped in the unnecessarytraffic jam but also some time in the regular condition they haveto wait for a long time-span even if the traffic density is veryless. This problem can be resolved if the traffic signal timer(TST) of the controller can be programmed to be manipulatedwith the continuously varying traffic density.is taken to minimize the congestion. This is the reason that oncethe congestion occurs, the more the time it will take to disperse.The more is the intensity of traffic burst the more will be thetime needed to decongest the area. So, to avoid the jammingcondition, the input rate, on one hand, is reduced by adjustingthe capacity of the cycle time movement's split, in accordancewith real-time congestion status, which on the other hand cutshort the incoming vehicle density (Jain et al., 2012).For preventing the traffic congestion by ensuring demandmanagement needs real-time traffic congestion status, in costeffective way. Using this traffic status, cycle time controlmethodology can be easily deployed to control the traffic lightsmartly. Various methodologies can be leveraged to get the realtime congestion data, some of these methods are comparativelyreliable but are more economically challenging, being resourceheavy, maintenance hungry and infrastructure intensive. Since adeveloping nation need not necessarily invest in high-levelcomputing infrastructure so a simpler method is described in thepresent research for real-time congestion resolution. Thetechnique involves the use of application program interfaces(APIs) provided by major companies in the field, like Google orMicrosoft, for real-time traffic status of traffic congestion. Thus,for optimizing congestion it eliminates the need for initialinfrastructure deployment for tracking vehicles and thus cutsdown the cost involved with the distribution of the sensornetwork. Apart from the low-cost sensing, this work also aimsat low-cost practical deployment of the proposed scheme with alight-weight application program (software) upgrade withoutthe need of new personalized full-scale hardware change.Hence, for sending the API sensed congestion deployment timeto traffic light master, ‘ZigBee’ communication boards are madeand tested that can easily integrate into existing traffic lightinfrastructure.3. REAL-TIME TRAFFIC CONGESTIONACKNOWLEDGMENT TECHNIQUES2. TRAFFIC CONGESTION COLLAPSE ANDCHALLENGESTheory of traffic flow explains that traffic congestion sets inquickly and takes a long time to dissipate once the networkoperates above critical point for very short time viz. 5 minutesonly (Elefteriadou, L., 2014; Knoop & Daamen, 2017). Trafficflow curve shows that the operational flow rate varies withtraffic density. For low density, the rate increases linearly withtraffic density and peaks at a maximum flow rate. After a peak,flow degrades exponentially with the increasing density tosaturate with maximum “road vehicles contain capacity”.Consider the case when the road link is operating at maximumflow for which they are designed, if a short burst of trafficenters the link and temporarily pushes the traffic density aboveoptimized value, then the flow rate will drop below maximum.This decreased exit rate will further increase traffic density ofroad link. This domino effect leads to the exit rate decayingrapidly that leads to congestion collapse and jam if no provisionThese jamming problems are more intense in developingcountries due to unplanned cities growth. However, to set-upextensive hardware-based traffic light real-time controllers andcongestion sensing infrastructure calls for significant amount ofinvestment which can scale-up with the increasing traffic. Also,in developing countries, various other hurdles are oftenencountered before the funds reach such demanding initiatives.Therefore, for practical deployment in developing countriesscenarios, a system with low investment is a sensible endeavour.3.1 Wireless Sensor Network (WSN)Sensor networks are the key to gathering the informationneeded by smart environment whenever needed. WSNs can beused to know the real-time traffic density around the area andhelp traffic light timing control system and as well as the driverto take several decisions in order to optimize arrival time andavoid traffic congestion. The sensor could be LIDAR orRADAR based sensor connected wirelessly or, induction loopand pulmonic tubes based infrastructure sending data wirelesslyto a center. Sometimes wiring is also used to communicate withcenter. The wireless sensors with integrated sensing, computing,and wireless communication capability make them easy toinstall but cost factor comes in the play when deployed for largearea. Currently, collecting traffic data for traffic planning andmanagement is mostly achieved by wired sensors in developednations around the globe. However, installation of wiredsensors and its maintenance cost limits large-scale deploymentfor monitoring real-time data that can be used to avoid thetraffic congestion and to look after the implementation of trafficrule (Pascale, A., et al., 2012).3.2 Video Surveillance and Camera FeedsYet another way to monitor the traffic-related problems is videosurveillance cameras (Jain et al., 2012). Much research has beendone to sense the traffic congestion through video surveillancecameras using low quality as well as high-quality video feedsfor both night and day. There are also some works leads tomerging information sensed from WSN as well as videosurveillance camera to give the more intelligent system to setfocus when congestion predicted (Collotta et al., 2014). Manyadaptive and fuzzy logic controller are based on surveillancecameras information collection. The major limitation of thisapproach is the wired communication infrastructure ofsurveillance camera involves considerable maintenance costs,high installation cost and reduces the architecture scalability.3.3 Infrastructure Free Vehicular NetworkRather than hardware sensing technique, infrastructure-freevehicular networks are reported, with very low adoption ofvehicle to vehicle (V2V) technology, at this time (Lu, N. et al.,2014; Hadded, Mohamed et al., 2014). Many works leverageThis contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full -83-2018 Authors 2018. CC BY 4.0 License.84

ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume IV-4/W7, 20183rd International Conference on Smart Data and Smart Cities, 4–5 October 2018, Delft, The NetherlandsVehicular Ad-Hoc NETwork (VANET) for various applications(Al-Sultan, S. et al., 2014). Also, several algorithms have beendeveloped to establish communication between vehicles usingVANET. Some algorithms used to cache data of congestedregions and pass the data to another vehicle are listed in Lakas,A., & Shaqfa, M., 2011. Due to non-prevalence of suchvehicles, these algorithms might fail to work sometimes,however to some extent they have addressed the problemsrelated to the data loss (Su, Hang, and Xi Zhang., 2007).Multichannel communications scheme to support audio/videoand other data are realized using Dedicated Short RangeCommunications (DSRC) for V2V applications. However, asmentoned, V2V is not so popular due to the high technologyneeded as well as sparse crowdsource data collection and roadswith limited V2V enabled vehicles, which may limit theapplication scenarios of the solution.intersection up to next consecutive traffic light in all thedirections as shown in Fig.1. However, the congestion buildingcurve (Jain, V. et al., (2012) suggests that for preventingchoking, congestion must not accumulate beyond a definedlevel for a given infrastructure. Therefore, in Google APIrequest, time input of ‘now’ is provided (the actual time ofrequest). So that it takes care of congestion from origin todestination from ‘now’ to up till the destination. Data resolutionprovided by Google API vary for different places by areasonable amount, say in less than a minute, and so can beconsidered as real time.3.4 Application Program Interface (API) AcknowledgmentThe vehicular topology is highly dynamic which raises somechallenges. The main challenges are the high mobility of nodes,intermittent links and stringent latency requirements. So,instead of applying VANET to forward the sensed andcalculated data, the sensed data using GPS and many othersensors in the mobile device are uploaded to the central serverwhere analysis and calculation are done. After, that theseprocessed data are made available to any client device orapplication in the form of APIs. Due to the advent of big data,machine learning, mobile computing and more technologies,these APIs are gaining many application usages (Ning,Zhaolong, et al., 2017).For retrieving estimated time to travel / arrival, there is an APIcalled Google Maps Distance Matrix API (Google Developers.,2017) that provides a matrix of origins and destinations fortravel distance and time. For calculating the ETA, ‘best guess’,traffic model of the API can be used. Which specifies theassumptions to use when calculating time in traffic. This settingaffects the value returned in the ‘duration in traffic’ field, inthe response which contains the best estimate of travel time,through historical traffic conditions and live traffic. Live trafficbecomes more important in case the ‘departure time’ is close tonow. Predictive travel time uses historical time-of-day and dayof-week traffic data to estimate travel times at a future date.This makes it easier to predict how long it will take to getsomewhere and therefore, is a promising parameter to measurecongestion.4. METHODOLOGY AND SYSTEM ARCHITECTUREThe proposed system adapts the traffic signal controlleraccording to the estimated traffic density in near real time. Aftersensing the congestion and mapping it in different congestionstatus level, the system sends the knowledge-based data ofdifferent cycle times available for 830 intersections all overDelhi (India) on (Delhitrafficpolice.nic.in., 2017). Along withsplit cycle time data available for Delhi and Bangaloreintersections, for congestion management, this whole systemworks on the problem by sending optimized cycle time to trafficcontrol system. Therefore, it is an endeavor to tackle thecongestion by adjusting demand management as it can’t beeliminated because the resources, i.e. road infrastructure islimited.The proposed method performs the congestion status check forthe point of deployment i.e. from the center point of roadFig.1 Highlighted traffic light under consideration is origin andCircled lights are destinationThe request trix/json?origins latitude,longitude&destinations latitude,longitude latitude,longitude latitude,longitude latitude,longitude&departure time now&key YOUR API KEY].Congestion Value (CV)CV (Max. hourlycongestion value of last week 2*Min. hourly congestionvalue of last week)/3(Max. hourly congestionvalue of last week 2*Min.hourly congestion value oflast week)/3 CV (2*Max.hourly congestion value oflast week Min. hourlycongestion value of MediumCongestion(2*Max. hourly congestion3Highvalue of last week Min.Congestionhourly congestion value oflast week)/3 CVTable 1. Congestion Status Lookup TableThe Google API responds through a JSON (data type) string tothis query by considering real time as well as near future trafficdata grabbed by their ML and AI algorithms using extrapolatingprevious records. The request can be made as many times as auser likes for a paid account, however, for testing purposes areasonable number of hits are allowed. Hence, for testing, APIresponse is called in every two minutes which is comparable tosmallest cycle time. Afterward, the status calculating algorithmdefines the congestion status in three different quantized levelaccording to the processed data grabbed by Google distancematrix API. The Google API response data provides maximumThis contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full -83-2018 Authors 2018. CC BY 4.0 License.85

ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume IV-4/W7, 20183rd International Conference on Smart Data and Smart Cities, 4–5 October 2018, Delft, The Netherlandsaverage time to travel between the origin and destinationconsidering the congestion on the way. For a fair quantizationand partitioning, three levels (1, 2 and 3) of traffic congestionstatus has been chosen as no or low congestion, mediumcongestion, and high congestion, respectively. As the differentknowledge-based data of each intersection for Delhi andBangalore is divided in 3 or 4 different cycle times according tocongestion hours, therefore, 3 levels of congestion are adoptedin order to avoid any quantization error. This API fetching andcongestion level calculation is done in authorized traffic servercentrally and the calculated cycle time is sent to local unit of theparticular site for deployment.Google API doesn’t publish its live congestion status dataexplicitly in digital form but mapping the response of estimatedtime to travel can be done. For this, s

Traffic light controller, Real-time traffic signaling, congestion, ZigBee communication board, Google Traffic API, Agent-based traffic modeling. ABSTRACT: Controlling of traffic signals optimally helps in avoiding traffic jams as vehicle volume density changes on temporally short and spatially small scales.