Braking news ! Predicting, pushing the limits and avoiding braking bad
The unconnected problem series, Story 4 —An efficient way to drive is to accelerate to a desired speed and then never touch the brake pedal. But are we connected to the world enough to achieve this ?
In nature, nothing is perfect and everything is perfect. Trees can be contorted, bent in weird ways, and they’re still beautiful. — Alice Walker
Could inertia be the unsung hero of efficiency ? An unlikely secret ingredient to make things more efficient while driving ? Since braking is going to remove energy and requires you to input extra energy to get back up to the desired speed, you’d get your best range by never braking ( under safe conditions ) in the first place. Regenerative braking is the next best thing.
According to the World Health Organization (WHO) report, road traffic injury represents the eight cause of death in the world and the leading cause of death for children and young adults aged 5–29 years. Every year the number of road deaths is around 1.3 million and between 20 and 50 million people sustain non-fatal injuries. Among the various causes, human errors seem to be the main cause of road accidents. In fact, speeding contributes up to 35% of fatal road crashes. Advent of new technologies like full autonomous driving at level 4 and 5 (AD) and its wider adaption could in future prevent such human errors saving lives.
As we move into the era of electric vehicles ( EV) and AD, the range of the EVs and the impact of AD on the energy consumption of the EV will be one of the key challenge. One study shows that AD capabilities could reduce the EV range with up to 30%. Could use of fast and reliable of innovative communication technologies such as 5G with connected autonomous vehicles improve the overall efficiency of EVs ? Early research suggests that 5G communication allows to overcome the limits of the commercial ACC systems like braking maneuver.
Assuming we know most of the environmental factors in advance then proactively slowing down ( without accelerating and utilizing the terrain or distance), is theoretically closer to an ideal scenario of optimal constant speed when compared to the current approach of relying on regenerative braking. Regenerative braking is a reactive way to redeem efficiency post braking event since the loss of kinetic energy is reclaimed by the generators to charge the EV battery with certain losses. So, efficiency gain with no braking should be more optimal than the energy recovery from regenerative braking under real world scenarios.
On long drives, especially on highways where environmental factors are less dynamic ( than in the cities ), currently autonomous vehicles can maintain speed and avoid obstacles, however they are still unconnected from a lot of environmental factors which determines the future need for braking hence reduces the energy efficiency. Let’s look at some of these factors and how efficiency could be improved in such scenarios.
Speed limits
Speed limits at certain segments of the road could trigger braking. Without advance visibility of the upcoming speed limit sections , an autonomous vehicle could try to maintain the speed by accelerating and expending energy before braking at the speed limit section of the road. Let’s visualize this scenario with the help of below image.
At point A , the speeding autonomous vehicle tries to maintain the speed over the incline by accelerating through to point B. As it approaches point C , due to the terrain the vehicle now gains speed and is over the speed limit. To compensate the vehicle must now brake hence resulting in loss of energy.
Advancements in artificial intelligence ( AI) and improved connectivity with 5G now opens up the possibility assess the impact of such static or dynamic local environmental parameters ahead of time and recommend the best course of action to prevent braking hence conserving energy. In this scenario local environmental factors like the existing speed, type of vehicle, terrain could have been considered to dynamically create a plan to slow down a few meters before the speed limit segment ( point A), such that braking is totally avoided when the vehicle arrives the the specific road segment ( point C). This is subtly different from existing AD mechanisms which will need the vehicle to apply brakes to slow down thereby loosing the efficiency. Let’s visualize the possibility with the image below.
In this scenario the autonomous vehicle is assumed to be connected in real-time via the 5G network slice, radio part of which is hosted along with the edge computing capabilities enabled at the nearest mobile tower. Apart from the real-time vehicle details like model, speed, the edge computer is also updated with the local environmental factors effecting the vehicle speed like road segment terrain, weather etc. One of the local environmental factor could be the speed limit at a certain segment of the road . When the vehicle is at point A, edge computer detects the vehicle speed at 83 mph and AI ( hosted by edge computing ) predicts the vehicle will exceed the speed limit at point D. The edge computation recommends a gradual slowing down by 3 mph ( without any further acceleration and leveraging the terrain). Vehicle further executes the recommendation so when it arrives at point D at time t+3, it achieves 80 mph without braking hence conserving energy. Though the overall energy saved in this instance may be very negligible, the idea is over a long drive such a scenario repeats itself multiple times and significant energy is conserved.
The toll queues
Presence of toll booths often slows down the traffic and during peak hours may even create long queues which requires the vehicle to brake and even idle in some scenarios. This is significant loss of energy hence reduction in range. If the local environmental factors also consider the situation at the toll like the wait time, speed, queue length and communicated in advance to the autonomous vehicle, the vehicle can now slow down ( by using terrain without acceleration) so when the vehicle appears at the toll junction there is no need to stop and idle whilst waiting in the queue.
The traffic lights
Sometimes the vehicles could accelerate at a specific stretch of the road only to stop at the traffic signal ahead when it turns red. Imagine if the AI at the connected edge computer could now predict the traffic light turning red at a certain time and accordingly relay the speed recommendation to the incoming autonomous vehicle, the vehicle could either speed up ( within limits ) or slow down ( without braking, using the terrain or distance) to conserve the energy.
Let’s take a look at an example where the autonomous vehicle without the advance knowledge of traffic light turning red at point B and time t+1.
The autonomous vehicle accelerates to 43 mph at point A and when it reaches point B, it encounters traffic light turning red which stays red for next 10 seconds. The vehicle now needs to brake and idle for next 10 seconds resulting in loss of energy.
Let’s contrast this with a better connected scenario where 5G slicing and edge computing capabilities are enabled and connected to the vehicle.
The connected vehicle now gets recommendations from the AI which predicts the time it takes for the local traffic signal to turn green. Based the time it takes for the vehicle to reach the signal at the current speed, AI then recommends the optimal speed required for the vehicle at time t ( point A). The autonomous vehicle works on this recommendation and instead of accelerating to maintain speed it now uses the terrain so it reaches optimal speed through point B ( time t+1 ) and reaches the signal at point C when it turns green after 10 seconds ( time t+2). At this point the vehicle no longer needs to brake and idle hence conserving the energy over time.
Road closures or traffic incidents
Advantage with 5G network is it can span over huge geographies hence virtually omnipresent. So if a certain terrain is not suitable to slow down sufficiently in advance, the neighboring cell tower could now be notified in advance to provide a early warning to the incoming autonomous vehicles of any recent road incidents or closures. Vehicles can now plan to slow down optimally with advantage of longer distance to do so instead of depending on the terrain to slow down.
The scenario is explained in the below image.
In this scenario as the autonomous vehicle approaches at point A the neighboring mobile tower or cell site will start communicating with the vehicle notifying of the road closure at point C and the optimal speed to maintain well in advance. This allows the vehicle to slow down gradually without using brakes , avoiding accelerating, so vehicle can approach point C at nearly crawl speed. In this scenario the energy is not expended at point A for any acceleration hence improving the overall range.
Tricky terrains
Another scenario could be tricky terrains which could play out in a similar manner as speed limit scenario and also include dynamic environmental factors like sleet or heavy rain conditions. The omnipresent nature of mobile network enables sufficient advance notifications to the autonomous vehicles based on such dynamic factors.
Navigating the winding connected path
I know nobody knows , Where it comes and where it goes — Steven Tyler, the lead singer Aerosmith.
The twists and turns of life’s journey can often be bewildering and unforeseen, leaving us with limited control and understanding over its many facets. Over the years, enhanced connectivity has demonstrated its inherent benefits for promoting efficient functioning within society. In an increasingly interconnected world, numerous unconnected points still persist, and their integration represents a crucial stride towards a more predictable future.
The convergence of connected autonomous vehicles with next-generation mobile networks like 5G emerges as the potential key to unlock an energy-efficient, safe and inclusive way of future travel, propelling us further towards sustainable progress.
