If you stacked every speed breaker in India, my estimate of about 20 lakh humps at the standard 10 cm height would make a tower roughly 200 km tall, twice the height at which space conventionally begins, with a plausible range of 66 to 660 km.
- 3.7 m × 10 cm is the IRC:99 standard hump, designed to be crossed at 25 km/h.
- ~20 lakh humps is a Fermi estimate (range 6.6–66 lakh); nobody publishes a real count.
- ~198 km tall, the stack would pass the 100 km Kármán line, then slump in under a second.
- 4.6 GPa at its base is over 100 times the strength of M40 concrete.
Calculated in this post from the 198 km stack height and my assumed 2,350 kg/m³ for dense bituminous mix (my estimate, no published figure), set against the 40 MPa characteristic strength of M40 concrete in IS 456:2000, Table 2: 4.6 GPa ÷ 40 MPa is about 115.
- 3,409 deaths were linked to speed breakers in 2015, about nine a day.
It would not stand: the bitumen mix that humps are made of flows under its own weight, and a column that tall would slump into a low mound within seconds. Melted down, the same humps would fill about 3.4 million cubic metres, around 1.3 Great Pyramids. Riding over all of them at sensible speeds would cost you about 53 days of extra riding time and roughly 5,000 litres of petrol.
Calculated in this post from the 65.93 lakh km road network in MoRTH's Basic Road Statistics 2020-2022, my assumed 0.3 humps per km (a Telangana study measured 1.66 per km on a hump-heavy urban stretch), the IRC:99 hump geometry and 2.6 million m³ for the Great Pyramid. The riding time and petrol use take the example inputs set out under the riding-time heading below.
No agency in India publishes a count of speed breakers, which is part of the joke and part of the problem. The Indian Roads Congress has a standard design for one; MoRTH has published how many crashes they cause; and anyone who has ridden NH16 knows they arrive unmarked, in clusters, at odd heights.
This what-if builds a count from data, piles the humps up, and asks the physics what happens next.
What does a standard Indian speed breaker look like?
In the early 1950s, the physicist Arthur Holly Compton, winner of the 1927 Nobel Prize for the Compton effect and then chancellor of Washington University in St. Louis, watched motorists speed past Brookings Hall on what is now Hoyt Drive. His answer was geometry. He designed what he called traffic control bumps, in a single-hump and a double-hump version, and they were first installed along Hoyt Drive in spring 1953; his notebook from June–July that year holds the drawings.
Newspaper clippings he kept suggest the humps were popular. The originals were later removed, and in 1992, for the centenary of his birth, two of his single-hump models were rebuilt, one at each end of the drive. The idea India would later pour by the lakh began as a physicist's sketch.
Washington University Libraries: Did Arthur Holly Compton invent the speed bump?The IRC:99 hump
The Indian standard speed breaker is a rounded hump 3.7 m wide and 0.10 m high, shaped as an arc of a 17 m radius circle, for an advisory crossing speed of 25 km/h. That design comes from IRC:99-1988, the Indian Roads Congress guidelines drafted from field studies by the University of Roorkee, and is the "hump" in every road engineer's head.
The same document says humps are meant for minor roads, residential streets and the minor arm of a junction, and that putting them on major inter-city roads "is not considered a good engineering practice". Where several are needed, it spaces them 100 to 120 m apart.
From that geometry we can get the volume of one hump. A slice through it is a circular segment of radius \( R = 17 \) m and height \( h = 0.10 \) m. Its area is
which matches the quick parabola shortcut \( \tfrac23 \times 3.7 \times 0.10 = 0.247 \) m². Across a standard 7 m two-lane carriageway that is about 1.72 m³ of bituminous mix per hump.
Calculated in this post from the IRC:99 hump size (3.7 m wide, 0.10 m high, 17 m radius) and a 7 m two-lane carriageway, the carriageway width used in the IRC:SP:73-2018 manual for two-laning: 0.246 m² × 7 m = 1.72 m³.
At an assumed 2,350 kg/m³, my estimate for dense bituminous mix, not a published figure, each hump weighs about 4 tonnes.
The guidelines define the standard hump (17 m radius, 3.7 m wide, 0.10 m high, 25 km/h advisory speed), allow 1.5 m long 1:20 ramps at each edge where trucks and buses are common, and set a spacing of 100–120 m centre to centre where humps are repeated.
They limit humps to minor roads, junction approaches, residential streets and similar places, require a warning sign 40 m ahead and black-and-white markings, and ask for before-and-after accident records. The limit: it is a 1988 "tentative" standard; IRC:99 was revised in 2018, and field surveys find most real humps do not match either version.
Read the standard (free)
How many speed breakers are there in India?
Nobody knows, so this is a Fermi estimate: about 20 lakh, with a defensible range of roughly 6.6 lakh to 66 lakh. A Fermi estimate, named after the physicist Enrico Fermi, breaks an unknowable number into pieces you can bound, multiplies them, and keeps track of the range.
- Road length. MoRTH puts India's road network at 65.93 lakh km on 31 March 2022 (provisional), counting national and state highways, district, rural and urban roads.
- Humps per kilometre. This is the uncertain piece. A study of a hump-heavy urban stretch in Telangana measured 1.66 humps per km. National highways should have very few, since IRC:99 advises against humps on inter-city roads, and most of the network is village and district road where humps cluster near schools, junctions and villages and are absent between them. I take 0.3 per km as a central value, 0.1 as a low case and 1 as a high case. That density is my assumption, not a measured national number.
- Multiply.
\[ N = L \times n = 65.93\times10^{5}\ \text{km} \times 0.3\ \text{km}^{-1} \approx 19.8\ \text{lakh humps} \]Low case 6.6 lakh, high case 66 lakh.
Treat the 20 lakh as a guide to orders of magnitude, not a census. City corporations and traffic police have counted humps on particular roads over the years, but no state or national body publishes a total.
The authors studied a road stretch with 1.66 speed humps per km using a test car mileage study, a graphical method to estimate mileage loss, regression of crossing speed against hump geometry, and VISSIM traffic simulation. The car used 68% more fuel on the humped stretch than on a free stretch, and the simulated level of service fell from A to E once humps were installed. The limit: one stretch and one test car, so the 68% describes a hump-dense urban road, not an average Indian road.
Read the paper (free)How tall would the stack be?
About 198 km at my central estimate, between 66 and 660 km across the range. Each hump adds its 0.10 m height:
For scale, the Kármán line, the conventional edge of space used by the FAI, is 100 km up. Even the low case, 66 km, is seven times the height of Everest, and the high case is more than six times the height of the Kármán line.
Use this to check what you just read: change the hump density, and watch the tower, the heap and the lost days move together.
Your own speed breaker tower
Why would the tower collapse?
Because the base of the stack would be squeezed far past what any road material can take, and bitumen flows. The stress at the bottom of a uniform column is just the weight of everything above it per unit area:
- Base stress of the 198 km stack.
\[ \sigma = \rho\,g\,H = 2{,}350 \times 9.81 \times 198{,}000 \approx 4.6\times10^{9}\ \text{Pa} = 4.6\ \text{GPa} \](I ignore the slight fall in gravity with height, which changes this by a few percent.)
- Compare with the strongest common material we could build it from. Even if every hump were recast as M40 structural concrete, with a characteristic compressive strength of 40 MPa, the tallest self-supporting column would be
\[ H_{max} = \frac{\sigma_c}{\rho\,g} = \frac{40\times10^{6}}{2{,}400 \times 9.81} \approx 1{,}700\ \text{m} \]That is under 1% of the stack. This is the same square-cube logic that limits the height of mountains: weight grows faster than the strength of the base.
Calculated in this post from the 40 MPa characteristic strength of M40 in IS 456:2000, Table 2 and an assumed 2,400 kg/m³ for concrete: 40×10⁶ ÷ (2,400 × 9.81) is about 1,700 m.
- Bitumen is worse: it is a liquid. The binder in asphalt is about 5–6% of the mix, and bitumen is a very slow fluid. The famous pitch drop experiment in Queensland measured pitch at about 230 billion times the viscosity of water, about \( 2.3\times10^{8} \) Pa·s. A viscous column relaxes on a time scale of roughly
\[ \tau \sim \frac{\eta}{\rho\,g\,H} = \frac{2.3\times10^{8}}{2{,}350 \times 9.81 \times 198{,}000} \approx 0.05\ \text{s} \]For a 1 km stack it is about 10 seconds. This order-of-magnitude estimate treats the whole mix as pitch and ignores the stone skeleton, which would make it behave more like a wet gravel heap; either way it does not stand.
What would you get instead of a tower?
A heap about 1.3 times the size of the Great Pyramid of Giza. The total volume is
Calculated in this post from the 19.8 lakh humps (the 65.93 lakh km network × my assumed 0.3 per km), the 0.246 m² hump section and 7 m width worked out above, and 2.6 million m³ for the Great Pyramid: 3.4 ÷ 2.6 is about 1.3.
| Pile | Volume | Mass |
|---|---|---|
| Great Pyramid of Giza | ~2.6 million m³ | ~6 million tonnes |
| India's humps, central estimate | ~3.4 million m³ | ~8 million tonnes at 2,350 kg/m³ |
Across the range of estimates it is somewhere between 0.4 and 4.4 Great Pyramids. That is a lot of bitumen spent on slowing traffic.

How long would it take to ride over every speed breaker?
About 53 extra days of riding at my central estimate, on top of the time to ride the roads themselves. The cost of a hump is not the hump; it is slowing down for it and speeding up again.
Assumptions, with these example inputs: you cruise at 50 km/h (13.9 m/s), slow to the IRC advisory 25 km/h (6.94 m/s) at 2 m/s², cross the 3.7 m hump, and accelerate back at 1.5 m/s². Bike plus rider 250 kg. These are my assumptions for a calm, legal crossing.
- Braking. Distance \( (13.9^2 - 6.94^2)/(2\times2) = 36.2 \) m, taking 3.47 s, against 2.60 s at cruise. Lost: 0.87 s.
- Crossing. 3.7 m at 6.94 m/s takes 0.53 s instead of 0.27 s. Lost: 0.27 s.
Calculated in this post from the 3.7 m hump width in IRC:99 and the example inputs above: 3.7 ÷ 6.94 = 0.53 s and 3.7 ÷ 13.9 = 0.27 s.
- Accelerating. Distance \( (13.9^2 - 6.94^2)/(2\times1.5) = 48.3 \) m, taking 4.63 s, against 3.47 s at cruise. Lost: 1.16 s.
- Total per hump and for India.
\[ \Delta t \approx 0.87 + 0.27 + 1.16 = 2.3\ \text{s}, \qquad 19.8\times10^{5} \times 2.3\ \text{s} \approx 4.6\times10^{6}\ \text{s} \approx 53\ \text{days} \]
- Fuel. Each slowdown throws away kinetic energy \( \tfrac12 m (v_1^2 - v_2^2) = \tfrac12 \times 250 \times (13.9^2 - 6.94^2) \approx 18 \) kJ. If the engine turns about 20% of petrol's 33.6 MJ/l into motion (my rough estimate), each hump costs about 2.7 ml, or about 5,300 litres for the lot.
Calculated in this post with these example inputs (a 250 kg bike and rider, 20% engine efficiency) and 33.6 MJ per litre for petrol: 18 kJ ÷ (0.2 × 33.6 MJ/l) is about 2.7 ml, and × 19.8 lakh is about 5,300 litres.
The cost of a hump is not the hump; it is slowing down for it and speeding up again.
If you instead built the 198 km stack as a spiral ramp at the IRC exceptional gradient of 6.7%, the ramp would be about 2,950 km long, longer than the ride from Bengaluru to Kolkata and a good way back, and your engine would run out of air long before the top.
Calculated in this post from the 198 km stack height and the 6.7% exceptional gradient in NPTEL's table of IRC gradients: 198 ÷ 0.067 is about 2,950 km.
On how far a rider can sensibly go in a day, see how many km per day on a motorcycle, and for long highway days, the 1,000 km day guide.
Do speed breakers actually cause crashes?
Yes, often enough that MoRTH used to count them: speed breakers were linked to 11,008 crashes and 2,633 deaths in 2014 and 11,084 crashes and 3,409 deaths in 2015, about nine deaths a day, according to the ministry's reply to Parliament reported in July 2017.
MoRTH's latest annual report, Road Accidents in India 2024, no longer lists speed breakers as a separate road feature; they fall under "Others". The categories it does list are below.
See the data as a table
| Persons killed, 2024 (deaths) | |
|---|---|
| Straight road | 118817 |
| Curved road | 20703 |
| Others (includes speed breakers) | 17510 |
| Bridge | 5982 |
| Ongoing road works | 5389 |
| Culvert | 4111 |
| Potholes | 2385 |
| Steep grade | 2278 |
Source: MoRTH Transport Research Wing: Road Accidents in India 2024, Table 3.5 · Total 1,77,175 deaths in 2024. Speed breakers are not reported separately.
| Year | Crashes linked to speed breakers | Deaths | Injuries | Source |
|---|---|---|---|---|
| 2014 | 11,008 | 2,633 | 9,428 | MoRTH reply to Parliament, 2017 |
| 2015 | 11,084 | 3,409 | 9,764 | MoRTH reply to Parliament, 2017 |
| 2024 | Not reported separately | Within "Others": 17,510 | - | Road Accidents in India 2024 |
For riders, the danger is the hump nobody marked. IRC:99 asks for a warning sign 40 m ahead and alternate black and white bands; many humps have neither.
RoadHook Rides' analysis of two-wheeler road deaths on Indian tours and the solo touring safety plan cover how to ride roads like this, especially after dark.

How sensitive are these numbers?
| Change | Count | Stack height | Extra riding time |
|---|---|---|---|
| Low density, 0.1 per km | 6.6 lakh | 66 km | 18 days |
| Central, 0.3 per km | 19.8 lakh | 198 km | 53 days |
| High density, 1 per km | 66 lakh | 659 km | 176 days |
| Central count, humps 15 cm tall (common in practice) | 19.8 lakh | 297 km | 53 days |
| Central count, harsher braking and 5 s lost per hump | 19.8 lakh | 198 km | 115 days |
The density per kilometre dominates everything. That is the honest weak point of this estimate, and the reason a national, geotagged hump survey would be worth doing.
You can collect India's speed breakers, but you cannot stack them. A 198 km column of bituminous mix would load its base with about 4.6 GPa, over a hundred times the strength of M40 concrete, and bitumen flows anyway, on a time scale of a fraction of a second at that height.
Calculated in this post as set out above, from the 198 km stack height, my assumed 2,350 kg/m³, and the 40 MPa strength of M40 concrete in IS 456:2000.
Recast in concrete, a free-standing column tops out near 1.7 km. The version that works is a heap: about 3.4 million m³, a Great Pyramid and a third, and you could ride up it on a spiral at a legal gradient. Keeping them off highways, as IRC:99 advised back in 1988, is the cheaper project.
What's new
- MoRTH's Road Accidents in India 2024: 1,77,175 deaths, about 20 every hour ThePrint (PTI), Jun 2026
- NHAI removes an unscientific speed breaker on the old Bengaluru–Mysuru highway; locals fear more speeding Public TV, Jan 2025
What if next?
- What if the Earth stopped spinning while you were on the highway? No speed breaker would slow you from 1,630 km/h.
- What if gravity doubled halfway through your ride? Every hump would hit your suspension twice as hard.
- What if you could ride a motorcycle across a lake? The speed you need to skim water, and why a hump is the last thing you want.
Questions people ask
What is the standard size of a speed breaker in India?
Under IRC:99, a rounded hump 3.7 m wide and 10 cm high, shaped as an arc of a 17 m radius, designed for a crossing speed of 25 km/h. Trucks and buses can get 1.5 m long 1:20 ramps at each edge. Many real humps are taller, narrower and unmarked.
Are speed breakers allowed on national highways?
IRC:99 says speed breakers are meant for minor roads and residential streets and that using them on major inter-city roads is not good engineering practice. On highways, rumble strips and signs are the usual tools, though illegal humps still appear near villages and junctions.
How many people die because of speed breakers in India?
MoRTH told Parliament that speed breakers were linked to 2,633 deaths in 2014 and 3,409 in 2015, about nine a day. Recent annual reports no longer list them separately.
How many speed breakers are there in India?
There is no official count. Multiplying the 65.93 lakh km road network by an assumed 0.1 to 1 hump per km gives roughly 6.6 lakh to 66 lakh, with about 20 lakh as a central guess.
Sources
- Washington University Libraries: Compton's traffic control bumps (first installed on Hoyt Drive in spring 1953; single and double humps; 1992 replicas; source for the story)
- Wikipedia: Speed bump (Compton's 1953 "traffic control bumps" at Washington University)
- Indian Roads Congress: IRC:99-1988, Tentative Guidelines on the Provision of Speed Breakers (source for 17 m radius, 3.7 m, 0.10 m, 25 km/h, 100–120 m spacing, 40 m sign)
- Korr RK et al.: Impact Analysis of Speed Humps on Indian Roads, Journal of EASTS 13 (2019; source for 1.66 humps per km and 68% more fuel)
- MoRTH Transport Research Wing: Road Accidents in India 2024 (source for 65.93 lakh km and deaths by road feature)
- Swarajya: Nine lives lost daily in accidents due to speed breakers, Parliament told (Jul 2017; 2014 and 2015 figures)
- Wikipedia: Pitch drop experiment (pitch about 230 billion times as viscous as water)
- Wikipedia: Asphalt concrete (binder 5–6% of the mix)
- Wikipedia: Great Pyramid of Giza (about 2.6 million m³, about 6 million tonnes)
- ThePrint: Road accidents in India claimed 1.77 lakh lives in 2024 (Jun 2026)
- Public TV: NHAI removes unscientific speed breaker on Bengaluru–Mysuru old highway (Jan 2025)
- Wikipedia: Kármán line (100 km, FAI definition)
- NPTEL: Vertical alignment I (IRC exceptional gradient 6.7%)
