[destinykid]

Originally Posted by ajayc123

As Mr Gadkari claims that the average waiting time at Toll Plazas has already come down to 47 seconds. What is the need to spend money to implement new technology to reduce this 47 s any further.

I believe the reason for this proposed implementation is not only just to reduce time as such, but also to deploy version two of this system wherein the toll collected will be based on the distance travelled. Presently, most of the toll gates charge full amount, irrespective of our entry and exist points. This camera based system might be provided at every entry and exit to reflect the exact toll amount based on the distance covered. Then again our existing highway system needs controlled access with well defined entry and exit ramps. We don't know what's on Gadkari's mind. But hopefully the idea would be to better the existing system. On the whole Fastag was a great deployment. But that doesn't mean that we shouldn't improve the shortfalls in current system and to achieve this, we need newer technologies and approach. Personally, I don't mind that working on better things as waste of public money. Innovation needs investment!

[thanixravindran]

Basically an ANPR camera recognizes a vehicle, take a picture of the number plate and then do a optical character recognition by enhancing the image. They need to recognize the number in the specified format only to detect a valid plate. Any fancy font, unrecognizable plate will obviously result in a error 'Invalid plates". Modern ANPR camera systems can identify the make and color of the vehicle also. They can also accept white listed and black listed numbers and can raise alarms and events like our CCTV. All of this is possible when the ANPR camera is integrated to a management system with database. For a stand alone ANPR camera, it is possible to achieve the above but with limited capacity.

ANPR cameras can work with IR or having a strobe light in less visible conditions. It is also possible to take a snapshot of the vehicle with driver (if tints are not used) once Invalid plates are detected, depending on country's legal position (privacy) and further action can be taken. In our case, my hunch is double toll will be charged for manual collection for invalid plate recognition similar to FasTag blacklist error. For politicos and muscle men, it is a free passage anyway. Given our number plate standardization project with HSRP is still stuck even after 20 years, it will be a challenge to implement. But knowing Govt, they will simply throw the ball to public to get HSRP if you want to avoid penal tolls

Biggest challenge for ANPR system in mass public system is the latency and bandwidth to scan multiple vehicles for the positive reinforcement cases (like toll) since most of the systems need latency of less than 250ms. This the reason it is used for negative cases of enforcement where only offending vehicles need to be captured (after meeting condition). Positive enforcement systems are typically used in private spaces like parking management at malls, office buildings, etc due to less load.

Now coming to the bank account linking, it will again mostly be wallet or recharge based since RBI mandates authentication for deduction of money from account unless standing instruction is provided at account level.

[PGA]

I feel its much odo about nothing. It's a system which has too many ifs and buts for it to be implemented effectively. It's one of those things that politicians speak at a heat of a moment. I will sleep over this news for next ten years atleast.

[vkivaturi]

Good to see Government departments exploring new tech to cut down on inefficiencies. ANPR, while not adopted widely yet, works well. Our apartment complex recently shifted to ANPR for boom barrier operation at main gate. Works fine in all weather conditions like low light, night time, rains. We earlier had a RFID based system.

Key benefits of ANPR (my personal view):

1. Avoids the need for additional sticker/RFID tag distribution
2. Captures moving vehicle number easily
3. Avoids the need for additional cameras for security and monitoring
4. Helps in removing toll booths completely as moving vehicle images can be captured, thus eliminating delays

Potential issues

1. Recognition of vehicle number may not be accurate in extreme weather conditions
2. Number plate duplication is a major risk. Not sure how this will be tackled.

[RaviDBLR]

Originally Posted by frewper

I dont see the point.
FastTag very clearly eliminated the time that was taken to pay cash.

The only thing that consumes time at a toll booth as of today is the slowing down of the car and wait for the barriers to open. Hence the only thing to eliminate is the toll booth itself.

Something flaky in the tech we have implemented that makes it slower to read? Singapore ERP and Houston EZTag I've used extensively work well with vehicle zipping through the toll booths.

Almost on every highway trip I seem to encounter at least one booth with the hand-held tag reader. The overhead readers does not work even when the car is stopped! Finally the handheld reader comes out to scan.

[ShivWason17]

Not fully aware about the technical details or pros & cons of this ANPR system vis-a-vis the fastag system but good luck reading a moving car’s number plate on Yamuna Expressway (or other expressways in north) in dense fog during winters.

[Contrapunto]

Originally Posted by RaviDBLR

Something flaky in the tech we have implemented that makes it slower to read? Singapore ERP and Houston EZTag I've used extensively work well with vehicle zipping through the toll booths.

Almost on every highway trip I seem to encounter at least one booth with the hand-held tag reader. The overhead readers does not work even when the car is stopped! Finally the handheld reader comes out to scan.

India we use a cheaper, passive "sticker type" tag and similarly more budget reader units. Understandable, since a very high of implementation would have killed it at birth. This can read at very slow speeds and at close proximity. That is why we need to slow down and possibly still need the boom barriers. I think the government also wanted to prevent lobbying and resistance from toll operators by giving signals that they are soon to become redundant.

What you see in Europe and US (and many other countries) are a more powerful tag that can be read from distant directional scanners even when the vehicle is at highway speeds. These look like a puck - similar in size to the Apple Watch charger. The camera simultaneously reads the license plate number and if you try to cheat by not paying the toll (not have a loaded account + tag installed), you get a fine and citation. This is what happens in a few rented vehicles and the customer is left with a high charge. Another advantage (as I understand it) is that it also has the chips to transmit accurate GPS coordinates so you can also be charged by distance since the last "Scan". As I understood from the scanty details and Nitin Gadkari's statements, the govt is looking at a system similar to this.

Now that the population at large is used to electronic tolls and the highway infrastructure is ready, it would be easier to roll out. All parts of the system would be expensive though- both for the car owners as well as for the scanning infrastructure itself. However, the toll booth operators have already downsized since they don't need cash collection processes and as may collection agents etc. They would be able to adapt, since their gross income is unlikely to go down.

[swunkjyn]

The system deployed should be smart, like if a number is associated with a vehicle, it should also recognize the vehicle by shape. So that if someone tries to use that number plate, since the vehicle is different, it should not let it pass.

[BrakeDancer]

Sounds like Mr Gadkari is inventing a solution for a problem that does not exist. Fast tags work just fine and a few seconds of waiting at the toll plaza to get it read doesn't seem all that to me, unless we're talking of emergency services. The direct linkage to bank accounts also sounds a bit risky and not something I'll personally be comfortable with. In this age of privacy concerns becoming more and more real with each passing day, this sounds like a step backwards. Technology that the government deploys is usually buggy, unreliable, and we can't rule out an incident tomorrow where because of faulty reading/scanning a person finds out that the bank account has been subjected to multiple deductions.

[autobahnjpr]

Many points to consider before we go further

• I am not comfortable with any system which directly links with bank account and does automatic debit. Even for UPI payments you need to enter the PIN each time to do a debit. Any system which authorizes direct debit to bank account is a risky proposition.

• Number plates are issued by each state RTO. You might have license plates not only in different shapes and sizes but in different languages. Till date we have not achieved standardization of license plates so any automation system which rides on top of it is designed to fail

• There are all sorts of vehicles including trucks, buses, tractors, mini buses, jeeps etc ply of roads and national highways. They all have their own way of putting number plate. Some put in center, some on the side etc. and some do not have it at all.

• License plates can get dirty and fade with time. Reading at the night will be another challenge.
  Also, I can put some other license plate and penalize that person as toll will be paid by that individual.

This look like another daydream and I hope nothing is done in this regard. Better we address issues with fastag and improve its working further.

[asmr]

Originally Posted by ajayc123

As Mr Gadkari claims that the average waiting time at Toll Plazas has already come down to 47 seconds. What is the need to spend money to implement new technology to reduce this 47 s any further

Since the question got raised, I am just thinking out aloud. May be removal of bottlenecks and the consequent issues that it creates? In heavy traffic areas or peak hours, the congestion at many toll gates is ridiculous. Cutting lines, fights with attendants, accidents are few things I can think of right off the bat as stuff that will disappear.

Our National Highways need to live up to their names by removing signals, barriers, crossings for people and vehicles by implementing clear, unobstructed exits, under or overpasses etc. Along with those, this removal of toll gates from main highways and/or moving to NPR without any bottleneck creation might actually cause less congestion.

I am actually in favor of keeping the toll gates as an access control mechanism OUTSIDE of the highways and not ON them. NICE road in Bengaluru and certain highways in Malaysia comes to mind as an example. I remember that the toll gates I passed through in Malaysia are mostly fully automated, much more faster in sensing the active "gun" that drivers hold in front of them with the "MyKad" (if I remember it right), pointing at the sensor.

[ajayc123]

I appreciate your perspective to suggest a barrier free automated system, that does not involve hunan intervention (Singapore ERP is one example). The figure of average of 47s seconds can be misleading without stating the peak hour waiting time.

On a standalone basis, the optical camera based number plate recognition is fraught with risks and would need to be complemented by other digital identifiers to ensure that accuracy is not compromised (like what happens in speed camera based solutions). The fasttag technology is definitely quite accurate.

[ritedhawan]

Lets agree that the current FASTtag system couldn't help decrease the time spent at toll plazas. There are a couple of reasons (some have already been outlined) behind it:

(1) The car when approaches the boom barrier needs to be aligned in that imaginary rectangle in order for the tag to be picked up. If a car ambles to the barrier, the driver needs assistance when to stop or continue ambling in order for the tag to be picked up by the reader device. This is where you find toll attendants waving their hands and then give you the go ahead. The time taken to complete this exercise should have been the only time spend by the car at toll booth before the boom barrier opens (add time spend in the queue as well). However, in some cases cars overstep the imaginary rectangle to the inconvenience of those behind the car. Because from that moment onwards that specific car will take additional time to get clear from the toll booth. The car that overstepped cannot reverse as these is a car behind it, thus the attendant walks up the car with a car reader and then the go ahead.
This adds to the inconvenience of those waiting in the queue.

(2) As already mentioned by someone in this thread, the system makes an API call in order to deduct the toll amount from Paytm or bank account and if the balance is insufficient then the toll barrier is not going to open. This is of particular inconvenience to those behind the car. Unless the car owner makes the payment via a different mode, the barrier wont open.

I think, instead of giving up on FASTtag , the government should make improvements to the current system. Before disowning it for an alternative.

[Mustang_Boss]

Gadkari sir proposed a GPS based system a year ago.

This would address all the issues with this proposed system.

The only concern is with the costs associated. If the Government could bear some cost for the old cars and get this implemented in the new Car's ECU or elsewhere, the issue would be addressed once and for ever.

I don't see a point in implementing a system where fellow citizens are penalized by others. I am not sure how efficient the resolution protocol going to be. So, implementing a fail proof, though expensive is the best choice.

[iamswift]

As most pointed out, now the payment bottleneck is removed, the only delaying entities are the reading of FASTag and the boom barrier.
Unless this is speeded up (efficient and fast reader, read-ahead etc), it will be just replacement of one system with another without significant improvement in toll booth time.

Another option is to have entry and exit tolls. Read on entry and calucate toll on exit. This requires toll reader in all possible exits but we can remove all the intermediate tools altogether. This can be an interim improvement until we have improve system to further reduce the time spent in tolls.