Planning safe roads


back to issue

ROAD traffic injuries and deaths have become a major public health concern in India with the total number of people involved in traffic crashes as well as fatalities per million persons increasing over the last three decades. Though at present non-motorized transport (NMT) and public transport trips constitute a vast majority of trips in Indian urban areas, the use of personal motorized vehicles (two-wheelers and cars) is rising along with increased risk to pedestrians and cyclists. This trend is accompanied with a rise in accidents and deteriorating air quality in cities.

There is a significant proportion of people who cannot afford personal motorized vehicles (cars and two-wheelers) for transportation and subsidized bus systems are also too expensive for them for their daily commute.1 They are dependent on NMT for travelling in cities. Even in the megacities of India (population more than eight million), more than 30% of the trips are made by NMT, a similar number by public transport (formal bus systems, informal bus systems and three-wheelers), and the rest by personal motorized vehicles (PMV), i.e., cars and two-wheelers.2

The poor quality of transport infrastructure and growing traffic congestion has been recognized by several expert groups and policy planners.3 At the city level, efforts to improve transport infrastructure since the 1980s have often involved road widening, junction improvement to facilitate movement of motorized vehicles, and construction of elevated roads. The specific needs of public transport users, bicycles and pedestrians have, however, not been included in the transport improvement projects. Any investment in infrastructure to improve mobility of motorized vehicles benefits only a small affluent class of people who own PMVs. Without facilities to regulate the interaction between motorized vehicles and NMT, this new infrastructure limits the freedom of movement of pedestrians and bicyclists substantially. Also, any investments made in infrastructure to improve mobility of PMVs result in increased vehicular speeds in the short term. This is often short lived, eventually resulting in an increase in congestion levels because of the increasing number of PMVs. Moreover, this results in increasing negative environmental impacts like deteriorating air quality, noise, habitat loss and fragmentation,4 and increasing number of accidents.5


Number of People Killed (in thousands) in Road Traffic Crashes in India Per Year

Source: NCRB, 2012.

Injuries are an important public health problem in India, contributing about 10% of total deaths in urban and rural India. In India, 137,000 deaths due to road traffic injuries (RTI) were recorded in 2011.6 This is among the three leading causes of death for people in the age group of five and 44 years. Nearly 15% of RTI deaths in the country occurred in cities with a population of more than a million. The rest of the deaths and injuries occur in districts and rural areas of the country, predominantly on state and national highways. The fatality rate has increased from 36 per million persons in 1980 to 95 fatalities per million persons in 2006.7


Fatalities Per Million Persons in Million Plus Cities, 2001 and 2006

Traffic fatality rate in Indian cities with population greater than one million. Source: NCRB, 2007.

Many cities show a fatality increase of 2-5% in recent years, regardless of the size of the city or the region. Small cities where newly upgraded highways have been built show the highest increase in fatality rates. The issues regarding traffic crashes in urban areas must be understood within a context that at present less than one in 40 families owns a car in India. The car ownership level is so low that even with reasonable economic growth (say 5-7% per year) most families are not likely to own a car by the year 2020. Consequently, a majority of our population is unlikely to use cars in the near future.


The data for types of road users killed are not available at the national or state level in India. Some cities maintain such details locally, but data are not available for all cities in the country. The proportion of road users killed in the late 1990s in the cities of Mumbai and Delhi, Kota, Vadodara and selected highway locations show that car occupants were a small proportion of the total fatalities (Figure 3). Pedestrians, bicyclists, and motorized two-wheeler riders accounted for 60-90% of all traffic fatalities. Children aged 14 years and younger comprise only 7% of the fatalities, though their share in the population is 32%. The proportion of fatalities in the age groups 15-29 and greater than 60 years is similar to their representation in the population, but the middle age groups 30-44 and 45-59 are over represented by about 70%.


Both land use policies and design of infrastructure have a major impact on RTI in cities. City planning policies that include the location of different activities, location of residential areas, and planning of transport networks influence the choice of transport modes as well as distances that various people have to travel. Mixed land use patterns reduce the length of trips and thus exposure to road traffic injuries. Often poor households are relocated to the outskirts of the city limits where land is cheaper. This results in long pedestrian and bicycle trips, and increasing exposure to road traffic crashes. Thus, road traffic risk to different road users is influenced by the city planning policies which decide where people can live and where the employment opportunities are located.


Road Traffic Crash Fatalities in Mumbai, Delhi,Kota and Vadodara

At the design level, design of road infrastructure (road cross section, carriageway width, intersection design), facilities for pedestrians, bicycles and public transport users influence the risk taking behaviour of road users. This includes observance of speed limits by car and bus drivers, waiting for sufficient gaps by pedestrians and use of zebra crossings and pedestrian subways.


This paper presents three case studies to show the impact of urban planning and road infrastructure design on the safety of road users. The first case study discusses the impact of relocating poor households from the self-planned locations in Delhi to the outskirts of the city for construction of the metro and other city development plans between 1997-2001. The second case study presents change in risk faced by pedestrians at a signalized intersection, which has been reconstructed as a signal free intersection to enable uninterrupted movement of vehicles. The third case study shows the impact of changing the existing road design for mixed traffic on a six km long corridor in Delhi to exclusive bus lanes, bicycle lanes and pedestrian paths on the safety of road users.


The last ten years have witnessed large-scale evictions and resettlement in Delhi. What lies behind the current spate of low income relocations are development projects like commercial complexes, flyovers, recreational parks, and wide roads to improve the landscape of the city. City planners have identified sites at the periphery of the city where poor households have been relocated. Peripheral development and relocation of urban squatters has meant an increase in the spatial segregation of social groups. This has also resulted in poor access to income generating activities.


Number of Households Moved Between 1997-2003

Site number

No. of households

Distance from original site



8 km



7 km



18 km



10 km



12 km



5 km



18 km



23 km



20 km



15 km



18 km



8 km



35 km



40 km



25 km

Source: Anand, 2007.

Table 1 shows number of households who have been shifted to locations planned by the experts. More than 40,000 households have been shifted from the central city location to the periphery of the city. This has resulted in an increase in travel distance to work as well as to the public transport stop.


We use indicators of accessibility and mobility to understand the change in RTI risk based on the change in travel distances as well as the mode of travel. Arora and Tiwari8 and Anand9 studied 2,000 households in Delhi to estimate the impact of relocation due to the metro construction in Delhi. The study documented accessibility and mobility conditions of households residing in the city in self-planned slums before the construction of the metro, and after they were relocated to new locations planned by the city authorities at the outskirts.

A. Anand estimated the indicators of mobility from the household surveys of low income settlements in the vicinity of the metro line and households who were relocated to new locations as per the land use policies to provide land for metro construction. The results from this study show that for the relocated households the value of all the indicators have changed. The distance to schools increased for 52% of the households but decreased for 41% of the households. Similarly, the distance to health services increased for 63% of households and decreased for 34% of households. Also, the distance to urban services increased for 52% of households and decreased for 36% of households. The highest impact is seen in the indicators measuring access to the bus system – the distance to the bus stop increased for 72% of households and the time gap between successive buses increased by more than 100% for 98% of households.


Interestingly, even for the households living in settlements which have not been relocated, there is some change in per capita trip rate (PCTR) for work (there is no change for 78% of households while it increased for 13%) and other purposes (there is no change for 82% of households and it decreased for 14%), but little change was seen in the PCTR for education (there is no change for 91% of households. The share of non-motorized vehicles (NMVs) in the modes used for travel does not change for 87% of households, increases for 7% and decreases for 5%. The distance to work, the time to work and the cost has not changed for 73%, 72% and 91% households respectively, and has increased for 17%, 17% and 5% households. For trips made for other purposes, the distance, time and cost indicators has not changed for 72%, 72% and 93% households and has decreased for 15%, 16% and 4% households respectively.

For a majority of the households relocated to new locations identified by city planners, the value of all the mobility indicators have changed. For 49% households, the PCTR for work has increased and for 30% it has decreased. For 71% of households, the PCTR for education does not change – it increases for 19% and decreases for 10%. The PCTR for other purpose has increased for 35% and decreased for the same percent of households. The share of NMVs in the mode used has decreased for 59% of households.

The mobility indicators for travel to work – distance, time and cost – have increased for 83%, 82% and 61% of the households respectively. The distance, time and cost of education did not change for 43%, 43% and 94% respectively and increased for 34%, 35% and 4% of households respectively. Regarding travel for other purposes, there is a decrease of distance and time for 58% and 52% households respectively but no change in cost for 65%.


The results of the study show that for the poor households which are not relocated to new areas, there is no significant impact on the indicators of mobility. The construction of a metro line does not change their mobility patterns. However, since the bus routes and location of bus stops were changed, these households face reduced access to transport services. With regard to the accessibility of households, while the land use accessibility remains unchanged, the transport accessibility has changed as distances to the bus stops increased for 19% and bus services became non-existent for 33% of the households.

On the other hand, poor households relocated to new areas experienced a significant impact on the indicators of accessibility and mobility. The land use accessibility has deteriorated as distance to schools, health services and other urban services have increased for 52%, 63% and 52% of the households respectively. The transport accessibility has deteriorated even more as distance to the bus stop has increased for 72% and the bus frequency has decreased, on an average, from five to 63 minutes (almost 13 times). The mobility of the households has also increased significantly. The PCTR for work has increased for 49% and decreased for 30%, implying a change in the number of trips made for work. The share of NMVs amongst the mode used decreased for 59% of households. The mobility indicators for travel to work – distance, time and cost – have increased for 83%, 82% and 61% respectively.


Per Cent Pedestrian Crossings and Accepted Gap at Signalized Junction (before construction of grade separated junction)

Gap of greater than four seconds denotes negligible risk.

It is well known that an increase in trip length by pedestrians and bicycles increases the probability of a fatal crash. Since the current planning policies have increased the distances of travel for households relocated to new areas, the risk of a fatal crash has increased. The mobility indicators for travel to work – distance, time and cost – have increased for 83%, 82% and 61% respectively. The relocated households are travelling longer distances than before on arterial or national highways coming to the city. These roads do not have dedicated facilities for pedestrian, bicycles or buses. These are highways coming into the city. Many households have been relocated along these highways. Therefore, the risk of being involved in road accidents increases for all families relocated as a result of our urban planning policies.


In Delhi, the government has made significant investments for the construction of grade separated intersections to make signal free junctions to reduce delays faced by motorized vehicles on arterial roads. With the construction of grade separators, pedestrian crossing problems arise. However, to facilitate pedestrian movement, pedestrian subways, i.e., underpasses and foot overbridges, i.e., overpasses have been constructed.

A study was undertaken at the All India Institute of Medical Sciences (AIIMS) flyover interchange in New Delhi.10 This intersection has large flows of bus, pedestrian, and motor traffic. The Ring Road, which is a major arterial road, and Aurobindo Marg form the AIIMS grade separated interchange. Traffic data collection allowed the study of road user behaviour both earlier, when the AIIMS junction was an at-grade, signalized intersection, and presently, when the site is a grade separated interchange with no traffic signal control. Our analysis produced results pertaining to pedestrian crossing behaviour as a function of observable pedestrian, environment, and traffic characteristics.


Before the reconstruction into a signal free grade separated interchange, the study revealed that 640 pedestrians used the southern cross-walk. From those, nearly 60% made a safe crossing (400 pedestrians did safe crossings, and 240 did partially safe or totally unsafe crossing). After reconstruction, 100% of the pedestrians observed crossed the road in an unsafe manner since there was no signal (344 pedestrians made unsafe crossings). Table 2 shows the approaching speed characteristics of the conflicting vehicles.


Speed Characteristics of Conflicting Vehicles

Vehicle group

Mean speed, km/h


Before reconstruction

After reconstruction


















Figure 4 shows the data for how people crossed the road; the percentage of all stage crossings versus accepted gap before reconstruction. It includes all unprotected pedestrian crossings for all stages, whether full or half. When the accepted gap is more than four seconds, the risk to the pedestrian becomes negligible. Totally, only 15% of pedestrians accepted high risks, that is, accepted gaps less than or equal to four seconds. The remaining 85% accepted negligible risk. Figure 5 shows the percentage of pedestrian crossings versus accepted gap after reconstruction. It includes all unprotected crossings that pedestrians completed. Only 38% accepted negligible risk. Figure 5 shows that accepted risk increased after reconstruction; more than 35% of crossings had accepted gaps less than one second as compared to 6% of stage crossings before reconstruction.

Previous research has shown that when the impact speed increases beyond 30 km/h, pedestrian fatality risk increases sharply.11 Table 2 shows the average speed of all motorized vehicle groups, which increased after reconstruction. It indicates that risk for pedestrians has increased. For instance, when the average speed of the car group was 26.5 km/h before reconstruction, the probability of death was approximately 6%. After reconstruction, the average speed of the car group increased to 32.5 km/h, thus increasing the probability of fatal crashes.


The study showed that a higher percentage of vehicles travelled at higher speeds in all categories after reconstruction. As a result, the risk to pedestrians increased. For pedestrians, the average accepted gap decreased after reconstruction in each stage of crossing, primarily because of the higher average speeds of the vehicle groups. The speeds increased 21.6%, 22.6%, 15%, 31.6 % for heavy vehicle, car, motorized three wheeler, and motorized two-wheeler groups, respectively. Twenty two per cent of pedestrians accepted shorter gaps (increased risk) despite the presence of a nearby pedestrian underpass. The study concludes that after the construction of the signal free grade separated junction, the risk to pedestrians increased substantially because the higher speeds of motorized vehicles forced them to accept shorter gaps for road crossing.


The Delhi government implemented a dedicated bus corridor in 2008. Buses on Delhi-BRT corridor operate in dedicated lanes, separated by a median in the middle of the road in an open BRT system. Because bus routes can join and leave the corridor at any point, passenger interchange time has not increased. Bus stops are at an average spacing of 500 m, mostly upstream of intersections setback by 20 m. Wherever intersection spacing is more than 500 m and points of significant boarding/alighting occur in between intersections, provision has been made for mid-block bus stops with signalized pedestrian crossing.


Per Cent Pedestrian Crossings and Accepted Gap at Signal Free Junction (after construction of grade separated junction)

Gap of more than five seconds denotes negligible risk.

The intersection design on the corridor aims to minimize conflicts and provide efficient passenger interchange. All traffic movement at the intersections is controlled through automatic signals. Traffic is segregated into bus lanes, motorized vehicle (MV) lanes and NMT lanes, each with their unique signal aspects, which may have overlapping or staggered green phases for different lane movements from the same approach.

Cyclists move on 2.5 m wide segregated lanes on the left side on both sides of the corridor. To reduce vehicular speeds, table top humps have been constructed at the entrance of cycle paths, and wherever a side road meets the main road to ensure the safety of cyclists. These lanes have been segregated from the MV lanes (in addition to 0.12 m wide and 0.15 m high kerb) by a 0.75 m wide median/unpaved zone on 75% of the length, more than 0.75 m wide green belt/footpath on 20% of the length, 0.3 m wide median on 4% of the length of the corridor.


Continuous footpaths are provided on both sides of the road that are wide enough to support existing pedestrian flows. At intersections, footpaths adjoin marked crossings for pedestrians – this maintains a continuous path for pedestrians. Pedestrian holding areas are provided at the kerbside, at each intersection, where pedestrians can wait before crossing the road. This area is also designed for street vendors. For pedestrian crossings, a 5 m wide zebra strip is designed across all intersection arms. This is preceded by a stop line (3 m away) to provide a safe zone for pedestrians to cross in front of the waiting vehicular traffic.12


A recent study has evaluated the impact of the new corridor design on traffic safety.13 The number of fatalities on the 5.8 km stretch of the Delhi-BRT corridor has been in the range of 4-17 per year with an average of nine fatalities/year in the period 2001-2006. November 2006-April 2008 was the construction period during which average fatalities were 6/year. In the first eight months of operation there were four fatalities. After further design interventions of controlling speeds by installing rumble strips in the bus lanes, two fatal crashes in the bus lanes were reported in 2009 (Figure 6).


Number of Road Traffic Fatalities Per Year (2001-2009) on the Delhi BRT Corridor (operationalized in April 2008)

As per the accident trends observed from 2001-2007 (Delhi Traffic Police records), it can be estimated that approximately 8.33 accidents were expected on this corridor if the BRT corridor was not made operational in 2008, i.e., the accident trend follows Poisson distribution.14 However, four accidents were observed on this corridor in 2008. This was further reduced to two accidents in 2009 after rumble strips were installed on the exclusive bus lane to check bus speed. This is to say that there has been a 43% reduction (standard deviation being 24%) in accidents after implementation of BRT and an overall 76.5% reduction (standard deviation being 17%) after installation of rumble strips on the corridor. The correction factor is 0.98 which is negligible. Traffic safety has increased on the corridor after it become operational.

The analysis also shows that of all the modes, safety for cyclists has improved the most as the bicycle interaction with buses on roads has reduced since the construction of segregated lanes for bicycles. Moreover, after the BRT corridor became operational, pedestrians faced the risk of impact by buses, but the implementation of rumble strip on the bus lanes resulted in reduced bus speeds, thereby reducing the risk imposed by buses to pedestrians. However, pedestrians continue to face risk by cars and two-wheelers, which needs further intervention to provide maximum safety on the corridor.


The three case studies discussed in this paper establish a very strong link between the safety of road users, policy and design interventions. Urban planning policies and land use policies decide the location of different activities and location of residential areas. Most of these policies have not been effective in addressing the needs of poor households who locate close to employment opportunities in the city, often squatting on the land not designated for residential use as per the master plan. However, this location results in short travel distances for pedestrians and bicyclists.


The case study from Delhi shows how the land use policies supporting the official Delhi master plan result in relocating poor households to the outskirts of the city to accommodate transport infrastructure like road expansion or the metro. This has increased travel distances and time for most households. The longer walking and bicycle trips on roads without any dedicated facilities for these modes increased the risk of getting involved in a fatal crash. The current relocation policies have thus made road users more vulnerable. The land use policies must ensure that the poor households, who cannot afford any form of motorized travel, are located close to employment opportunities, thereby reducing travel distances. This will bring down the risk of fatal crashes because of reduced distances and travel time, in addition to increasing accessibility to education, health facilities and employment opportunities.

The other two case studies discuss the impact of road design interventions on the risk of fatal crashes. Unfortunately, the conventional understanding of measuring performance of a transport infrastructure is biased towards car traffic. Thus the level of service of an intersection is measured in terms of delays faced by motorized traffic. Unsurprisingly, this has become a major source of concern for planners, traffic policy and road owning agencies. Road expansion schemes and signal free junctions have become synonymous with ‘improvement’ of transport infrastructure.

Since problems faced by pedestrians and bicyclists, the two most vulnerable road users, are not viewed as major transport issues, the ‘improvement’ strategies do not take into account impacts on pedestrian movement. The conversion of a signalized junction at AIIMS in Delhi to a signal free junction, for instance, has resulted in an increase in motorized traffic and increase in risk faced by pedestrians while crossing the road. However, when road designs factor in the needs of pedestrians, bicyclists and public transport vehicles as the Delhi BRT case study presented, the number of crashes can be reduced.

The case study also shows the impact of design interventions on the speed of buses. The exclusive bus corridor resulted in high speeds and involvement of buses in traffic crashes. However, after the installation of rumble strips in the exclusive bus corridor, the number of bus-pedestrian crashes reduced. Road designs which explicitly address the needs of bicyclists and pedestrians, and ensure speed control, have a major impact on road accidents.



1. D. Mohan and G. Tiwari, ‘Mobility, Environment and Safety in Megacities: Dealing with a Complex Future’, IATSS Research 24(1), 2000, pp. 39-46.

2. RITES, Traffic and Transportation Policies and Strategies in Urban Areas in India. Ministry of Urban Affairs and Employment, Government of India, New Delhi, 1998.

3. MOUD, National Urban Transport Policy. Ministry of Urban Development, Government India, New Delhi, 2005.

4. H. Demirel, et al., ‘Exploring Impacts of Road Transportation on Environment: A Spatial Approach’, Desalination 226(1-3), 2008, pp. 279-288.

5. M. Peden, R. Scurfield, D. Sleet, D. Mohan, A.A. Hyder, E. Jarawan and M. Colin, World Report on Road Traffic Injury Prevention. World Health Organization, Geneva, 2004.

6. NCRB, Accidental Deaths and Suicides in India – 2011. National Crime Records Bureau, Ministry of Home Affairs, New Delhi, 2012.

7. D. Mohan, O.T. Simhoni, M. Sivak and M.J. Flannagan, Road Safety in India: Challenges and Opportunities. UMTRI-2009-1. The University of Michigan Transportation Research Institute, Ann Arbor, MI, 2009.

8. A. Arora and G. Tiwari, A Handbook for Socio-economic Impact Assessment (SEIA) of Future Urban Transport (FUT). Transportation Research and Injury Prevention Programme (TRIPP), Indian Institute of Technology Delhi, 2007.

9. A. Anand, Socioeconomic Impact Assessment (SEIA) Methodology for Urban Transport Projects: Impact of Delhi Metro on the Urban Poor. Ph.D. thesis, Indian Institute of Technology Delhi, 2007.

10. U. Gupta, N. Chatterjee, G. Tiwari and J. Fazio, ‘Case Study of Pedestrian Risk Behaviour and Survival Analysis’, Journal of the Eastern Asia Society for Transportation Studies 8, 2010, pp. 2123-2139.

11. E. Pasanen, Ajonopeudet ja jalankulkijan turvallisuus [Driving speeds and pedestrian safety]. Teknil-linen korkeakoulu, Liikennetekniikka, Espoo, 1991.

12. RITES and TRIPP, Operating Plan for HCBS Corridor – Ambedkar Nagar to Delhi Gate. Report for DIMTS. Delhi Integrated Multi Modal Transport System Company, Delhi, 2006; TRIPP, First Delhi BRT Corridor: A Design Summary – Ambedkar Nagar to Delhi Gate. TRIPP, Indian Institute of Technology Delhi, 2005.

13. G. Tiwari and D. Jain, ‘Accessibility and Safety Indicators for All Road Users: Case Study of Delhi BRT’, Journal of Transport Geography 22, 2012, pp. 87-95.

14. E. Hauer, ‘The Naive Before-After Study’, in E. Hauer, Observational Before-After Studies in Road Safety. Second ed., Pergamon, 2002, pp. 73-93.