The Curious Case of VKM Measurement

Why regular travel data are crucial

Road transport in India is one of the key pillars for the country’s economic growth. It accounts for about 87% of passenger traffic and 60% of freight traffic movement in the country. It also accounts for 12% of India’s energy-related carbon dioxide emissions and is a major source of criteria air pollutants.[1] The increasing demand for private motorised mobility and the transport of goods are likely to strain the existing road infrastructure, drive up energy consumption, and consequently transport-related emissions in a business-as-usual scenario. With the emergence of new energy vehicles driven by batteries or hydrogen, there is a rising need to invest in a very different format of support infrastructure in the form of charging and hydrogen refilling stations. Motorised road transport in India is growing and evolving and so must transport planning. This warrants meaningful analysis based on credible and granular transport data to produce insights that help in effective policymaking and planning.

Vehicle Kilometres (VKM) forms the empirical basis to carry out a wide gamut of analyses with several applications.

Vehicle Kilometres, an important parameter to track in detail

Obviously, when one talks about data, it essentially means monitoring and studying specific parameters that possibly helps understand the scenario. Distance travelled by vehicles is a critical information that leads to a parameter usually referred to as Vehicle Kilometres (VKM) or Vehicle Kilometres Travelled (VKT). It is the total distance travelled by motor vehicles in a given period. Although its meaning sounds simple, it is a versatile factor. It forms the empirical basis to carry out a wide gamut of analyses with several applications. These include but are not limited to:

• Profiling vehicle-stock based on the degree of road usage. This helps, for example, identify the set of vehicles in terms of vehicle form factor, use-case or type of registration that currently utilises or requires road infrastructure more than others.
• Understanding the energy or motor fuel consumption pattern by type of vehicle that can possibly be linked to contribution to emissions.
• Getting a sense of the economic attractiveness of transitioning to new-age vehicle technologies like electric drivetrain as the cost-effectiveness of adoption depends on VKM travelled.
• Planning new-age road transport infrastructure like charging stations and associated upstream electricity network and supply.

By studying multi-year VKM statistics, one would be able to decipher the change in pattern across all these issues. While analysing the data, it is paramount to take cognizance of the fact that motorised road transport has significant heterogeneity and so has VKM statistics.

It is paramount to take cognizance of the fact that motorised road transport has significant heterogeneity and so has VKM statistics.

Therefore, it is critical to collect and analyse the VKM-related data with due consideration of the diversity (in terms of vehicle form factor, use-case, registration, etc.) in the vehicle stock. There can be even other factors for divergence in the VKM value like the size of an urban agglomeration or the type of motor fuel used (since there are differences in fuel prices). Are the VKM statistics able to reflect such heterogeneities? This depends on the techniques employed to estimate.

Methods for VKM estimation

No direct measurement of annual VKM travelled has ever been tried; the available information consists of estimates based on various assumptions.

No direct measurement of annual VKM travelled has ever been tried; the available information consists of estimates based on various assumptions. The accuracy of the assumptions and the sampled data influence the reliability of the estimates. Historically, the importance of evaluating VKM travelled has been from the perspective of roadways planning and other non-energy-related areas. Understanding traffic density on key road stretches has been the focus.

There are two main conventional methods used to understand travel-distance data: the vehicle-count method and the fuel-consumption method. Both are indirect ways.[2]

Vehicle count method is the most widely used technique for estimating VKM travelled. This procedure assumes that VKM travelled during a year can be estimated by counting the traffic on a representative section of a roadway, usually between two major intersections, during short periods of time and extrapolating these results to the entire geography, be it state or country. The reliability of this monitoring is based on the equipment used, as well as on the location and the frequency of the counts. Equipment is extremely costly to use on an extensive network of roads. Moreover, the traffic flow differs significantly between urban, peri-urban, and rural roads.

Fuel-consumption method assumes that vehicle kilometres travelled in a year is a function of the total volume of fuel consumed by vehicles during the given year and the average fuel economy of vehicles. The simplicity of the method has some inherent flaws. On one hand, the data for fuel consumption by vehicles is not captured by type of vehicle. The origin of this data is usually the aggregate fuel sales figures reported by the fuel-refilling stations.[3] On the other hand, the average fuel economy values although can be disaggregated by vehicle form factor are usually based on lab tested conditions which differ from on-road vehicle performance.[4]

New-age VKM measurement technique

In addition to the conventional approaches, there is a potential new-age technique thanks to the digital footprint of human activities including travel. The ubiquitous mobile smart phones have turned out to be a novel and useful source of collecting granular travel data. The very feature of mobile smart phones, enabled by GPS[5], to check real-time vehicular road traffic on a map during travel can be utilised to get a sense of the travel pattern over a geography. From the pattern, VKM travelled by an individual all the way up to the population can be derived. However, there can be multiple challenges due to high uncertainty in mobile phones-generated distance travelled data. It is difficult to attribute GPS-based individual travel data to kilometres covered by a vehicle.[6] It is even more challenging to filter the data by vehicle form factor which effectively limits the usefulness of the derived VKM data. There are also concerns that use of real-time digital tracking of individual activity can be considered invasive in nature and hence, any analysis based on such data may face regulatory and public scrutiny.

Need for an out-of-box solution

While there are multiple ways to estimate VKM travelled, more often than not the captured VKM-data suffers from issues like questionable quality, expensive or time-taking process, less data-granularity restricting the scope for meaningful insights, over reliance on assumptions and extrapolations, and limited opportunity for time-series analysis. Also, the data available lack clear reference to the underlying measurement method which raises doubts about the validity of the data.

Considering these shortcomings, there is a need to look for alternative ways that produce more reliable data on VKM travelled in Indian context and are easy and cost-effective to employ. Why not leverage an existing monitoring system in India’s road transport sector?

A regular touchpoint between motor vehicles on road and the state machinery are the vehicle pollution testing centres that are commonly co-located with fuel refilling stations dotting major urban agglomerations across most parts of the country. As mandated by the Central Motor Vehicle Rules (CMVR), 1989, every motor vehicle has to come to an authorised pollution testing centre for periodic testing and obtaining PUC (Pollution Under Check) Certificate which usually has a validity up to 6 months.[7] With the current digital integration of the pollution data collection system with a national database managed by the Ministry of Road Transport and Highways (MoRTH), it makes vehicle data recording and archiving easier. Apart from capturing pollutants-related parameters, a PUC Certificate has data fields concerning the vehicle such as registration number, date of registration, fuel type, etc.

Clearly the existing system of vehicle pollution testing at the PUC Certificate issuing centres can be piggybacked on to collect other useful vehicle data including VKM travelled.

The existing system of vehicle pollution testing at the PUC Certificate issuing centres can be piggybacked on to collect other useful vehicle data including VKM travelled.

How to record VKM data using the vehicle testing system

Every motor vehicle has an odometer that displays the total distance travelled by the vehicle since its purchase. This value is essentially the VKM travelled by the motor vehicle. The odometer reading (Figure 1) can be easily recorded using visual, photographic, or videographic method.

Further, currently it is mandatory to video record the testing process and upload the footage on the “PARIVAHAN” portal.[8]  This gives the opportunity to record the odometer reading of a vehicle when it comes for testing.[9] Adding this new data field essentially does not require a different step, a new technique or technology or re-training of the personnel at the pollution testing centres. The personnel at the testing centre just have to take the reading and feed to the new data field online that will subsequently get populated on the existing national database of MoRTH.

To formalise the VKM data collection exercise, an amendment of the existing testing norm allowing the recording of the odometer reading would suffice. This needs to be complemented by awareness-building through mass media about the new requirement.

How can the odometer readings be used to measure annual VKM?

The reading taken every time the vehicle goes for testing generates time-series data from which VKM travelled by the vehicle during the time intervals can be derived and in a span of a year i.e., after two consecutive 6-monthly testing, the annual VKM for the vehicle can be calculated. Linking with other information of the vehicle already captured, the VKM data thus collated can be analysed by filtering with different factors like vehicle type, registration, fuel, vehicle age (based on registration date), and geography (depending on the registering state or the location of the testing centres).

To allay any concern of the vehicle owner, the amendment should categorically mention that the collected odometer data will be available on PARIVAHAN portal at an aggregate level and not at an individual vehicle level. Also, the captured reading should not appear on the issued PUC Certificate to prevent individual-level attribution.

Key advantages

This ingenuous way of collating VKM data offers several advantages over the existing conventional methods. The proposed process:

1. Ensures collection of far better quality of VKM data as it is a direct method and does not involve assumptions or low-confidence extrapolations.
2. Allows logical segregation (filtering) of the data, thus capturing the heterogeneity of the road transport sector.
3. Makes available updated data as the database will be a living repository. This helps track changes in mobility pattern.
4. Provides a wide geographic coverage at state and national levels.
5. Is simple and cost-effective to execute.
6. Does not infringe into vehicle-owners’ privacy and data confidentiality.

Is there a possible challenge in the process? The suggested system would be effective as long as the pollution testing centres remain relevant in India. This implies that only a scenario when the vehicle stock (not sales) is dominated by new energy vehicles may potentially make the proposed data collection system obsolete. For India, this possibility is still far-fetched and distant.

Additional details

Fuel-consumption method

This method assumes that vehicle kilometres travelled in a year is a function of the total volume of fuel consumed by vehicles during the given year and the average fuel economy of vehicles. This means:

FC = Total fuel consumption by vehicles in a geography during a year;

FE = Average fuel economy of vehicles.

The advantage of this fuel-consumption based procedure is it neither requires any expensive, complicated instrument for monitoring nor involves a complex analytical method to derive the result. Both the fuel consumption data and the average vehicle fuel economy values at a broad level can be collated with relative ease. However, the simplicity of the method has some inherent flaws.

Vehicle count method

To express the estimation logic mathematically, as per this method total VKM travelled during a year is:

where

C_i,j = traffic volume (count) passing location i during period j;

L_i = length of the representative section of the roadway where i is located;

w_i = assigned weight (or expansion factor) to equate L_i with the total set of roads it represents;

t_i,j = assigned weight (or expansion factor) to equate the count during period j with the total annual count at location i.

There is no standardised sampling procedure to apply this technique.