India has hadseen a staggering increase in its population within the past 30 years. Couple thepopulation increase with the economic growth of the country and we arrive at atotal of over 20 million cars registered by the end of 2017. Due to the trendof urban migration of the workforce a high percentage of these cars drive inthe urban centers of India. If we have a high concentration of cars in arelatively small area of land, in this case a city, we soon realize that the emissionsof these cars are a major contributing factor to the problem of air quality. In order tokeep the current population and future generations healthy, we need to act nowand prevent further pollution of the environment. The following proposal is anattempt to act with currently available technology to manage the air pollutionin major urban centers.
By no means, is this a solution which will solve theproblem we as a developing society face, but rather an attempt to control thedamage done and to prevent it from extending further. The main ideabehind this proposal is continuous monitoring of air quality all over the city.Sensors would be installed all over the city. Each sensor would then transmitits data to a central station where the data will be analyzed. Since airpollution cannot be modeled as a linear system, many more environmentalvariables need to be take into consideration such as, air humidity, wind speedand direction, temperature, precipitation etc. Once the allthe data is collected we fuse it together to estimate the status and obtain a value,which is also a dynamic function and updates real-time. This value is thenautomatically compared to a pre-calculated threshold value.
If the threshold isexceeded, an alarm is set off and the authorities are notified. According tomagnitude of the overshoot of the threshold value, a certain percentage of carswould need to be withheld form the roads for a specific period of time. For instance,the population would be notified two days in advance that for a period of twoweeks each day only cars with license plates containing randomly generatedalphanumerical combinations are allowed to drive. This approachis a novel one of course and would need the full cooperation of the citizens towork. This however in reality most certainly would not be the case. For this reason,CCTV cameras mounted on stoplights at intersections would continuously monitorthe license plates of cars during the restrictions. Anyone caught driving a carwith license plates which do not contain the allowed alphanumerical combinationwould be issued a fine, which would be automatically added to the property tax.
Repeat offenders would see an exponential growth of their fines. The restrictionwould only apply to personal vehicles, and not to public transport. On thesedays, public transportation can be supplemented with additional vehicles. Onceenough data has been gathered, the problem can be modeled as a machine learningtask, namely as a regression problem. The would enable the algorithm to predictfuture air quality values, giving the authorities more time to react.