Musing on the future of mass transit.

Transporation in the 21st century.


The major problem with adding any new transportation system is cost. Cost is typically driven by size. Therefore, the first step in creating a better transportation system is to shrink the vehicles.

Imagine a vehicle half the width of a bus/train. Trains and buses typically have an isle and four seats. This new vehicle just has two slightly larger seats. This enables two vehicles to run side by side in the same carriageway or for two-way traffic to run on a traditional single lane width.

Secondly, reduce the height of the vehicle to that of a car and use separate doors for each seat. A traditional car door design could work or they could slide. The point is to make a vehicle more like a car, ones you sit in rather than stand up in and walk around.

The benefits of this are obvious. A new dual-direction transportation system needs only half the land area of a traditional system. For underground transport, only a single narrow tunnel can serve both directions. This is important because the cost of tunnels is a function of their diameter.

For highly congested areas the duel carriageway can be split into a fast and slow lane. The vehicles would only enter the slow lane to pick up and drop off passengers. The fast lane is that – fast. Having a dedicated slow lane means the vehicle can stop where the passenger wants, not where the stops are. But also it means that the when stopping it won’t be holding up the vehicle behind. This is a common issue that trains and buses encounter.

In comparison to a bus or train, the vehicle is low and narrow. Creating a very small frontal area, reducing aerodynamic drag and improving efficiency. The vehicles would also be far lighter than a traditional bus. Owing to their reduced drag and weight, their energy consumption would be less. And acceleration and de-acceleration would be improved.

Automation and Machine Learning

The vehicles would be automated. Readers might think “We don’t have that technology yet.” However, many public transportations around the world are fully automated. As long as the vehicles are segregated from normal traffic full automation is achievable now.

Automation allows the driver to be removed. This eliminates human error and reduces costs. Here we should remember why buses and trains are large, it’s because they need to transport the greatest number of people with minimal drivers. With automation, this is no longer a problem.

Automation means that many smaller vehicles could run instead of fewer larger ones. These small vehicles can be programmed to move to where they are needed. Because they are faster than traditional buses, and not held up by traffic or other vehicles, their effective capacity is increased.

Furthermore, because this is an entirely new electronic automated transportation system new technologies can be used to improve capacity. Instead of using a traditional system of providing vehicles on set timetables, they could come when requested, like an Uber. The vehicles not in use could then dock and charge at strategic locations and charge or simply not waste energy. There is little point of having many vehicles moving around when there is little demand for them. When someone needs to travel they go to a stop and press a button to request a pickup or they request a pickup on their smart-phone.

Admittedly this sounds similar to Uber, but it would operate on specific networks for example on commuter routes.

With machine learning, the network can be trained each day to know where each passenger is travelling to.

Over time the network will learn the best way to organise the passengers. Instead of each passenger getting on a random vehicle, they will be designated a vehicle to travel on. This vehicle will select passengers travelling to locations to minimise journey time. It may pick up more passengers on the route, or it might simply drop off all it’s passengers, switch onto the return route and return to its original departure location to pick up passengers again.

The benefit of machine learning with smartphone data would be that humans wouldn’t need to try and guess what capacity is needed at a certain time. The machine knows and each day the network would gain more information about how many passengers need to be picked and where they are going. Then the fleet can rapidly be adapted on the fly to remove bottlenecks. This is contrary to buses, and trains, which run on fixed timetables.

Other Advantages

No large infrastructure for stations required. The vehicles are light the carriageways need minimal reinforcement reducing cost. Maintenance of vehicles simplified because of their small size and relatively few components. No ticketing system as stops are paid for by smart-phones or contactless cards. Option to run vehicles on normal roads as automation improves enabling more destinations to be reached. Elimination of NOx emissions. Much faster transit times than buses increasing popularity.

The relatively low cost of this system could enable countries with little space for road expansion to convert current infrastructure like bus lanes with minimal costs. Alternatively, old closed railway lines deemed too expensive to reopen could use this cheaper form of transportation.

The Future.

The future is electric, it is autonomous. If we are to improve public transport around the world it requires looking at it with a new perspective. I firmly believe the answer to that is to shrink the vehicles. It is time to stop thinking big. It’s time to think small.