Tsinghua Open Courses | From intelligent vehicles to smart transportation

okay uh 720 let's get started okay uh very good evening

everyone it is a great pleasure to speak in front of our young bright audience uh my name is shab I'm a

professor uh with school of V call and Mobility at chinua University at Beijing China I used to teach in Australia for

six years from 2012 to 2018 and I I taught in Sweden Europe for

four years from 2018 to uh 2021 before returning to my home country in December

2021 uh my research area is focused on Next Generation transportation system uh today I'm very delighted to

present with you the future Transportation Systems highlighting a few Trends and driving forces using a

few key studies that my group had been doing in the past five six years uh this is a very exciting area and we need

young bright bring to join us for further development in this era in transport and the vehicle uh this is my email address if

you have you are very interested in my topic uh send me an email and we can communicate okay uh I will start with

the economic map of two major country on Earth uh who Le leaders just met with

each other yesterday in B Indonesia uh after not seeing each other for very long period uh as it can be seen in the two

diagram see the two diagram uh 1.8% land uh in the United States contribute half of their uh

GDP and 0.23% of land in China contribute half of our

GDP uh another country with massive land Australia 7.7 million square kilometers

is 45% people and 65% economy uh in two cities in Sydney

and Melbourne two major cities in that country uh the reason I mention this uh I use this diagram to

demonstrate to emphasize that urbanization you see the first sentence urbanization is Unstoppable is the

Unstoppable Trend and there are huge economic culture uh political educational and health

benefit U however urbanization comes with cost uh the most important the most difficult one is traffic congestion

which substantially affect the livability of a city you know and the wellbeing of the residents living in the

city so we don't like traffic congestion and even worse even worse the rapid uh urbanization

amplifiers the most fundamental problem of Transport the most fundamental problem of Transport which is the asymmetry between

Transportation demand and Transportation Supply uh I will use these two diagram two videos to explain uh the left hand

side is transportation demand uh the right hand side is Transportation Supply for Beijing Subway system uh left hand side for

transportation demand we always have very heavy traffic on the morning PE hours right on the morning PE hour from

7 o' to 9:30 very very crowded very very heavy heavy tra traffic while the volume at night is very low volume at night is

very low the demand is changing that means the demand is changing dynamically from time from time to time

the transportation demand is different on the morning Peak is very crowded on the night low demand period only

few users are using our transport infrastructure in contrast the supply in

the right hand is relatively stable uh why because the supply is determined by by what by infrastructure right by by

the transport infrastructure if you have transport Transportation infrastructure you have the capacity for transportation infrastructure it it the capacity is the

fixed value it is not wise for us for the designers to use the highest demand that means the morning Peak to design

our infrastructure as the it will be a huge risk we usually use 75 percenti 75

percentile that means under 25% of the time we are congested 75% of time no problem however with the rabit

urbanization we experience contests at many more occasions uh for example in Beijing the conenction period you know we are living in Beijing conenction

period uh is not limited certainly not limited to the 25% of time uh these are two uh examples these

are two examples uh during the P or you see left hand side the Pik or in Tokyo Japan it's very crowded you see the prin

sters have to push the paers into the train while during the off Peck hour uh here uh even in Mumbai India the train

is almost empty right so the PE hour very crowded ofp or no traffic no

users and you know in the past several years in Beijing during the P hour in in some femer train station is not the

train is full of people or full of users is the whole train station occupied by users by their residence

commuters so after explaining this background uh in the following slides I will introduce a few key studies and I will use this key study to demonstrate

what will happen in future uh do this few key studies that were done by my group in the past several years and I want to share with you the latest

progress in the area of transportation and the vehicle okay when we consider when we consider the traffic congestion problem

we will need to deal with the having Bon right so for any problem any issues we first to find the Bon to find the

critical point and what are the most common traffic bon in the section right in sections and we only have partial

time you know a certain percentage less than 40% even less than 35% for One Direction the time is the Green time that can be used to to discharge the

traffic so take a look at this video take a look at this video I will introduce what we can do to solve the problem the problems

in the video with connected automated vehicles in our first key stud so I will introduce uh you know three large Kee

studies in tonight in today's lecture uh okay take look at the video first

okay then the next and then you only intersection you always watch unfold the can you hear the

voice you can yes fundamental problem of traffic on green the first car accelerates and then the next and then the next and

okay then the next and then you only to catch the red had the cars accelerated simultaneously you would have made it through coordination not cars is the

problem because we are monkey drivers with slow reaction times and short attention spans even if we tried getting everyone to press the pedal on 3 2 1 now would be challenging this

discoordination limits how many cars can get through an intersection and when one backs up to the next that's when city-sized gridlock Cascades happen taking forever to clear in general more intersections equals more

discoordination which equals more traffic this is the most behind big highways no intersections splits and merges yes intersections no no stopping no coordination problems no traffic well

that's the theory anyway intersections outside of a highway will back up onto it again because human reaction times limit how many cars can escape the off-ramp when the light changes but even without intersections there would still

be traffic on the highway traffic can just appear take a on Lane Highway with happy cars flowing until a chicken crosses the road the driver who sees it breaks a little the driver behind him doesn't notice no immediately and breaks

a little harder than necessary the driver behind him does the same until someone comes to a complete stop and oh look cars approaching at highway speeds must now stop as well though the chicken

is long gone it left a phantom intersection on the highway this is what's happened when you're stuck in traffic for hours thinking there must be a deadly pile up ahead and then suddenly the traffic's over with no wreckage in

sight to you a relief if you're a good person and Mild annoyance if you aren't you just pass through a phantom intersection the cause of which is long gone and this Phantom intersection moves

it's really a traffic snake slithering down the road eating oncoming cars at one end and pooping them out the other on a ring Road a single car slowing down will start an AR roboris of traffic that

will last forever even though there's no problem with the road if the drivers could coordinate to accelerate and separate simultaneously easy driving would return but they can't so traffic

Eternal on Highways traffic snakes grow if cars are Eed faster than excreted and they shrink if excreted faster than eat dying when the last car accelerates away before the next car must stop now in

multi-lane highways there needs be no chicken to start gridlock a driver Crossing Lanes quickly with cars too close behind is enough to birth a traffic snake that lives for hours and leagues it's this quick Crossing that

causes drivers behind to overbreak and begin a chain reaction but we can make traffic snakes less likely by changing the way we drive your goal as a driver is to stay the same distance from the

car ahead as from the car behind at all times tailgating is trouble not just because it makes accidents more likely but because you as the tailgator can start a traffic snake if the driver

ahead breaks always in the middle this gives you the most time to prevent over breaking but also gives the driver behind you the most time as well and when stuck in traffic this rule would

get all cars to pull apart the snake faster that's the simple solution to traffic getting humans to change their behavior perhaps by sharing this video to show how and why traffic happens why

Tailgators are troubled and how we can work together to make the roads better for all the end except yeah Wishing Upon a Star that people are better than they

are is a terrible solution every time instead what work okay in the following I will introduce

that what we can do to to resolve the problem with self-driving Vehicles

okay works is okay so for human drivers as mentioned just now we have limitations we have reaction time we have slow reaction time

we are not cooperating with each other we made mistakes right it seems that not

you know impossible or not very likely to solve the for solve this problem if if we are we are the driver we are the

human drivers and the Very origin of the problem is is all own limitation right it's very unlikely that people's re reaction time can be shortened right

it's not very likely it's not it's unlikely that our perception range you know the perception range can be

enlarged however with the emerging in ICT uh information and Computing technology and vehicle technology we are

in the process of fundamentally resolving problem through connected automated Vehicles which will be our first casye

study as I mentioned just now uh we propose uh modernize the Transat that will be on second you know later I will

introduce second key study and we propos the low altitude vertical Transportation which I will introduce in the key study

three okay the first key study it to use connect automated vehicles or connected self driving vehicles to Dien the

oscillations or to solve the problem just now the the snake problem just me mentioned in that video and we want to

get rid of the Stop and good activities or academically the oscillation stop and go activities as I mentioned earlier uh

with human beings with human drivers we have our own limit uh we have reaction time the drivers are different from one another and your your reaction time is

different from your neighbor's reaction time and we are unwilling to cooperate or even if we have will cooperate we

don't know how exactly we can Co cooperate to maximize the throughput uh these are the reasons that

about mention problems in order to resolve the problem we as vehicle Engineers we develop the connected aut connected vehicles and automated

vehicles uh mainly as metion that's now attributed to the recent development and progress of information and Computing

technology the vehicle vehicle technology and very recently the AI Technologies the connected automated vehicles don't they don't have reaction

time right they don't have reaction time they are robot and they have much larer perception range than human driver who

can only rely on our own eyes right the connected automated Vehicles they have the driving range much larger than human

drivers however for a single connect automated vehicle the enhancement the Improvement in efficiency is rather Limited

what if we have many more Vehicles which are connected automated and Cooperative what we are having okay let's take a look at this uh

two videos uh the upper video you see the uper video uh the xaxis is the time is time uh you see the red signal green signal red signal green signal red

signal green signal red signal right the time and the y axis is the location is the location the position of the vehicle

right so for each line represent the trajectory of a vehicle the time position trajectory of each vehicle

right okay let's take a look at the video again at the normal dri at the normal Drive uh in the traditional interfection which is the upper video we will follow

the speed liit right you are normal driver you will follow the speed liit the speed is the slope of the trajectory right you follow the speed limit to

approach the infection when you see the red signal you see the red signal and you you will decelerate right you will decelerate and stop in front of the

stopping line and wait for the green signal to discharge when we see the green signal what will happen what you will do what will you do

you will have two seconds of reaction time and Accelerate from speed of zero to speed of speed limit to pass through

the intersection right that's the whole process you decelerate stop wait you see the green signal you have reaction time

after the reaction time you Accelerate from zero and then you want to discharge the intersection so here you have you take

look at the upper video you have the you have three problem right the B neck is in the in however I in the infection you

have the lowest lowest efficiency you have the worst efficiency because you have the reaction time here two second

GRE time visted you have stop and go activities in the in the B neck stop and go activities that means you have high

energy consumption you have lower speed you have more uh probability of making mistake and you have uh safety consist

if you have stop and do activities because people are different people are different from one another and also here the speed average speed is very low

right very low speed if you have very low speed that means when you discharge the interaction you take longer time to distar in interfection so no good no

good the system is no good what what should we do what should we do you see the lower video the lower video you you focus on the lower video what with the

connect automated Vehicles what we can do is to decelerate in advance you see this vehicle decelerate in advance and then you accelerate by the time when you

arrive at the infection uh you see the said at the very beginning of the green signal you use the speed highest speed to discharge to pass through the

interfection then you see the capacity here is much higher the capacity here is much higher so you use the highest speed to

discharge the infection and there is no reaction time here there is no reaction time and people are following each other to discharge the to pass through the

infection that means you have no stop and good activities you have less probability of making mistake if you check the overall performance you have

the TR travel time you have the reduction of travel time you have the reduction of energy consumption because you avoid stop and good

activities and you have the higher throughput you have higher cap capacity for each and every second of your green

time your green time is better utilized and in terms of safety is much better as well because here you have stop and good

activities you might have mistakes if you have mistake you have big mistake you have C but here you don't have stop and go activity you people are easier it

is easier for people for drivers to follow each other to pass through the inspection then stop and good activities so you have TR to benefit you have

energy consumption benefit you have safety benefit for this kind of uh if all you have the uh connect automated

vehicles uh this is another comparison uh when we consider two directions you know the two direction two traffic flows

comparing comparing with uh no control scenario uh especially for the Q lens and delay you know much better for

all uh connect automated Vehicles much better with the correct with the correct control

approach okay then you may ask now we only have a very small fraction of vehicles that are connected automated

right you don't have 100% And the previous example you don't have 100% connect automated vehicles what we can

do what we can do if we you take a look at the uh subfig b subfig b only 10% connect automated vehicle that means out

of 30 Vehicles you only have three connected connect automated vehicles on this

case uh we will not be able to control each and every Vehicles right we can but we can still regulate

these three vles right um we can still uh regulate the trajectories of these three vehicles and influence other

vehicles to have a collective benefit okay and we can regulate the vehicles for these three connected automated

vehicles and then we regulate other vehicles so we com we we influence other vehicles so this examples shows the

platoon of 30 vehicles uh a platoon of 30 vehicles that means the vehicles are successively arriving one

by one with a very similar time dis difference from one vehicle to another vehicle in other words the arriving

pattern is uniform in uniform arrival in the left hand side all the vehicles are traditional vehicles that

means 0% 0% connect automated vehicles and you have the Stop and good activities you have the illation

right while the right hand side with 10% these three vehicles are connect to the vehicles these three vehicles become the

leading vehicles of the three different platoons you see the leading Vehicles following following following by followed by seven

traditional Vehicles the human human manual Vehicles human driven vehicles and the second green time you have the leading

vehicle the platoon of seven human driven vehicles and the third one is also okay so that means you have three platon

and you have three platon and we can minimize the STS and enhance the safety efficiency and Energy savings so you see uh for this scenario you only have three

stops right the first three vehicles see see we have to stop for all the other vehicles they only have fast slow and fast

we they don't have stop and go activities right as it can be seen in the this diagram the sub

subfigure uh you know C D and F referring to the travel time the energy consumption and safety and we can make

our transport system safer you know faster and green for each and every Vehicles almost Collective effect is

very substantial especially for safety you see the safety the difference the difference Gap is very

substantial okay uh when it comes to the multi L multi flow system you have left hand vehicle you have through vehicle

you have right hand vehicle so when when we when it comes to the multi Lan multi flow problem uh the Boton neck is always

the left tening problem so in the right right driving system such as in China in the left tning system in Australia in UK

in Singapore uh the problem is the right hand problem right is the Big 10 is the Big 10 problem for the through Vehicles you see take a look at this this diagram this

diagram the first one for the through Vehicles you have three typic L right you have three through lens if you have green signal all the vehicles in these

three through lens can discharge right simultaneous they can discharge together so the efficiency is is high

but you see the left handing problem you see the left handing problem the left tning Vehicles you only have one left tning L and the opposite Direction you

also have one left handing line one left handing line one left handing line if you give the signal to those left

Vehicles the green signal this two lens this one line and this one L this the opposite direction left tanging Vehicles these two lens blocks the whole traffic

because all the other vehicles they are not allowed to distra because otherwise they will be conflict right so that

means this green time the utilization is very very low because you only have one L the vehicles in one L can discharge in

this green time for left 10 Vehicles so this is the problem we want to resol this is the problem we want to rol so

now if we have connected automated Vehicles we can reorganize the tempic flow in real time to separate the left

tening vehicles and the through Vehicles so take look at here you given any combinations you given any

combinations you the uh the black cars left handing Vehicles the white Cs and through vehicles through a vehicle sorting method approach we can

reorganize the traffic flow we organize the traffic flow we put the stre left t Vehicles ahead of the three through

Vehicles then we give the left handing signal and all three lens can be used to for the left handing vle purpose right

so the the the efficiency can be triple increase two times and then we give the green signal to the three vehicles all

three lens can can be used to uh to discharge take a look at this video so we have any combination and for this

combination we are using four steps that means less than 6.5 seconds so we reorganize the P flow all the left handing vehicles are ahead of the

through Vehicles the left 10 Vehicles you give the left hand signal and all three L can be used for left hand purpose and then we give the through

through uh green signal for through Vehicles all the through three LS can be used for the for the through vehicles e

okay just now I will not allow to unmute myself so now you can see my voice right okay

uh now we finish the multi system multi system so another problem uh another problem is about uh free about free you

may notice that you know when you are driving on the 3v or you are riding on the 3v you may notice that there are actually uh for this example there are

actually three typic lens in this example but only two of which can be used for Mobility purpose right the other one is always reserved for

emergency vehicles right emergency vehicles however 90 over 95% of time nobody is using the emergency L nobody

is using the emergency l so in the past with the uh tradition Vehicles there is little we can do there is little we can

do uh we have to ensure that uh the absolute priority for emergency vehicles uh that means for ambulance for police

car and for fire right okay in the past with the traditional Vehicles there nothing we can do we construct three lens but we

can only use two lens right is the V of infrastructure capacity nowadays with the connect automated Vehicles we can reorganize the traffic flow uh we can

reorganize the traffic flow I mentioned just now if there is no emergency vehicle in the system all the lens we use all the three lens all the lens can

be used if there is any emergency vehicle coming over we can reorganize the traffic flow to clear the 50 meters

in front of emergency vehicle guarantee the absolute priority for the emergency vehicles guarantee the speed of the

emergency vehicle we will have a much better system with not only good for the social Vehicles because for social Vehicles you you only have two lens uh

but now you have three lens can be used but also for emergency vehicles because for emergency vehicles we guarantee there will be no any vehicles in the 50

meters ahead of the emergency vehicle guarantee is absolute priority and is speed uh these are some uh simulation

results and that's what done by my group if any you know fire engine coming over or ambulance police car and they will be

given the priority and make sure the spe will not be affected and arrive at the incident site as soon as

possible ambulance arrive at the incident side as soon as [Music] possible okay uh that's the end of the

first case study and that's the case study one for The Connect automated vehicles and now uh let's shift or

to the second key study a new Transportation mode a new public transport mode and Moniz modularize tens

uh for most commuters in large city for most commuters in large city the public transport is always the first option it

is cheap it is green it is sustainable right however we have a lot of complaints right we have a lot of we are not happy with public transport uh we

have a lot of complaints first of all you see the the complaint number one it

stops at each and every station it stop at each and every station even if only

one passenger allies or BS the whole bus had to stop the rest users will have wied their travel time

so for for a normal bus with 50 people you know for a normal bus with 50 people so if only one pattern VES to get off

the rest 49 passengers the travel time of rest 49 passengers will increase because they have stop and accelerate

they have this activity which is no good right so the first problem we don't want to have stop at each and every station the second problem is inconvenient

transfer inconvenient transfer because uh roughly in Beijing uh for the other trips uh 39 38 no over 35% trips will

need transfer to complete to finalize their uh their trip their origin to destination for each transfer you have

to stop and you might need to change platform so for some platform so for some station you have numerous St platforms and you need to wait right you

need to wait for the connecting bus so on on one hand you know others uh and and we mentioned just now this the de TR

time on the other hand your o power time the uncertainty will be increased significantly right you also have to risk you have you

also have the risk uh of meeting your Comm connecting bu and you know you don't know the the route from one one platform to another platform you take

long time than necessary and then you miss your connecting bu is very likely so this is the second problem

inconvenient transfer the third problem is the Ved capacity uh you always have crowded you you are always crowded

during the pick but however during the nighttime even if only one passenger on

the whole route the whole the whole h bus has to operate uh it is very visible when it comes to energy consumption the

O all operating cost for the whole tree for the whole rout you only have one penger but you need to the the bus the

Big Bus need to oper so that is very beastful so how to deal with this problem how to deal with this problem

using the traditional basses uh it is unlikely to resolve the problem uh but with the you know connect automated vehicle

technology uh you know we it is likely that two two cabins can be connected to couple or

decouple by emotion so one of our collab collaborating partner designs uh a models taking advantage of the recent

progress of information Computing control connected automated vehicle technology the vehicle the they call it

port or cabin the cabin can CLE or de couple while motion while all those

cabins are moving so the passengers can move from one cabin or one port to another cabin while em motion so then

this system can resolve the three problem that I mentioned just now during the morning PE or we can have multiple cabins linking together maximize the

utilization of Transportation infrastructure maximize the efficiency while if somebody first problem take look at back the first

problem is stop at each and every station right uh if you have six cabins linking together you know during the morning PE if one passenger want to Al

light see we can for all the lighning passengers they move they transfer to the last cabin and then we get rid of the last cabin to for lighting and

boarding purpose and for all the other passengers we don't need to stop we don't need to waste the travel time right so we solve the problem one the

problem two and the in com in the transfer as I mentioned just now as it can be seen in the video the passengers can be transferred can be transferred

from one one cab to another cab right while in motion and then so for the transferring cabin you can get the rid

of the C transfer in cabin and the transferring cabin can link together with the uh the vehicles in another

Direction just like this video this vehicle they can go to another route linking together with other vehicles so

then solve the second problem of trans uh of inconvenient transfer the third problem is uh the during the off peak

hour during the nighttime is very of in terms of the energy in terms of the energy consumption in terms of the uh

the the operating cost so for this problem if you during the morning you have six seven Vehicles linking together but during the night you only have one

vehicle you know in the in service all the other vehicles go back to the depot for charging right for charging so if the demand is low the most of cabin most

Vehicles can go back to the depot for charging so they are not they don't need to operate so in some what we can do is

that during the P the multiple Vehicles can be connected and reduce the congestion because the road utilization

rate is higher so for vehicles between you know one vehicle and the subsequent vehicle you have two second time

difference that means this 2 seconds if the speed is 16 m per square per second that means the distance between the

first vehicle and subsequent vehicle is 30 plus meters but with this system all the vehicles are linking together the infrastructure utilization rate the

density much higher right uh during the off PE or a small number of uh vehicles uh can stifi can sany the demand so

improving the Energy Efficiency as well as the service col quality because during the off PE or it can operate like taxes

right so my group has done a series of product in the vehicle LEL because we have the uh connect we have this kind of

uh Next Generation buses and uh and also the connecting pass uh we also have done a lot of work in the operational level

in the design level in the planing level uh we have already uh have you know the vehicles uh and we are now keeping we we

are now keeping to improve it so we aim to replace many taxes because taxes or taxi you may have one

driver and V right is not the efficiency is not high the car occupancy is not high and we wish to occupy a great

portion of the ma share so in our team we are also using the discrete Choice model to estimate the market share of

this new Mo uh competing with the buses traditional buses competing with the taxes the new new market share for buses

for cars for this new mod will be different than they are today uh this is a very interesting

extension uh you know this is a very interesting extension as all vehicles are connected or physically connected they are physically connected uh we can

certainly realize the interch charging right interch charging from one can to another C interch charging from one vehicle to another vehicle in this

regard we can have our a small local energy cloud system namely if you need some energy uh if you are L of energy

and you can buy or borrow some energies from your NE neighboring Vehicles so you link with your neighboring Vehicles interch charging and then you get energy

from your NE neighboring vehicles with this with this we will no longer to have uh the range anxiety right we will no

longer to have a red anxiet we we also don't need to have huge battery consider your own driving

Behavior Uh your your your REM remaindering battery are always greater than the kilometers you want to travel

for a given day or at least is a very very high probability your reminding battery is greater than your kilometers

you want to travel uh that means you have the reserve the energy for seal right for uh you know for leav for SE uh

to or to to borrow the energy to your uh neighboring Vehicles considering the scenario that's you know another

scenario or violence charging technology have made break throughs so we can have Val Charing with 70%

efficiency while all vehicles are emotion vehicles and another vehicle they have the different the distance gap

of 20 M or 25 Metter you have 70% efficiency then we will have energy cloud system for the whole city right instead of a local energy cloud system

system you will have energy cloud system for the whole city for the whole city on this case the enies can transmitted from

one vehicle to another vehicle VI wireless charging and then the energy can be allocated based on the demand based on the need right so now the

question is in Europe in Europe I attended uh quite a few uh meetings in this area they wish to establish 65

million charging times before 2030 uh they want to establish 25 million Charing PS before 2030 and

consider the establishment as the precondition consider the establishment of this 65 million charging packs as the

precondition for transportation electrification if we have this kind of interch charging system uh we are either

modelization physically connected and commit and the energy transmission all VI charging uh not physically connected but VI charging has made break through

the energy can can be transmitted through wireless charging from one vehicle to another vehicle VI the establish uh establishment of the

charging time still be precondition for the transportation electrification so on this topic you see

the diagram you see this diagram uh for on this topic we have established a model uh we have two findings you know

today I'm I'm not going to uh introduce the technical details today I just introduce some findings some performance

uh some benefit that using connect automated vehicles and using emerging uh transport mode using emerging uh

Technologies uh the two findings number one we don't need to have so many Charing PS right you see you don't have we don't need to have so many charging PS we only

need to have less than one third of char impact number two our battery size can be significantly smaller than it is today so we don't we don't need to have

huge uh battery because if we have the range problem we can always borrow energy from our neighboring vehicles uh

if so we we have to consider you know the technological development of charging

devices either wireless charging or physically charging we need to consider the charging uh Technologies with the

charging infrastructure planning because if you on 2030 you have 65 million charging P but you made the Breakthrough

you have the uh vience charging from one vehle to another vehicle so what will happen you need to you need to uh uh get

rid of these charging parts right you need to remove all the charging parts that's a vist of the construction

maintenance all the cost and the removal cost okay that is the end of uh the second key study uh we are not going to

have any uh class break and I will finish the pr and hopefully we will have around 10 to 15 minutes for the um you know questions

and answers uh the key studies three is the low altitude vertical transportation system so this is the vision uh for uh

that proposed by my group for the low altitude vertical transportion system uh this figer draw by myou uh this see the

in Beijing you see the CCTV you have the drones for the logistics for the puzzles uh you have the flying vehicles uh you

have the modular buses model cars uh you have high speeded real right uh the drones and you as as you can imagine drones can be used for

Logistics purpose right for the puzzle for delivering puzzle from one area to another area certainly we need to examine you know the problems like noise

right you have the Privacy issue privacy issue and you have the regulation ISS right but we can uh but we are certain that this issues can be resol can be

resolved uh let's consider the current the parel Deliverance uh in your own country the puzle Deliverance uh it's more or less like this kind of system a

track loads puzzles from all the customers at a warehous or Depot with with this

a sequence of distribution points so you see this is the circle you read it a number of distribution points or

distribution centers and the return the warehouse the return the warehouse so you see that this the truck and then for this distribution point you will have at

each distribution Point uh you will have a lot of fillar uh usually in Beijing is T

tricycles or three whe Vehicles tricycles that can easily access the residential area uh they are responsible

for the last segment delivery for each Traer uh we travel from the distribution Point deliver the passel to the uh to

the end users and return to the uh Distribution Center this is how most of

our current current service work uh how all at least how we expect they to work uh in the United States uh they are

also examine the possibilities of using drones for the last segment so they they

are also using the TRU uh to for the uh for the rout from the wareh to Distribution Center distribution center

and go back to the warehouse but the their trucks are equipped with uh drones uh and the charging facilities U

and all the passengers in the warehouse and you know the trucks very similar to this this this scenario the trucks will

go through the rout and then you have the distribution Point use the drones to replace the pillers to replace the traillers and for the last uh for the

last segment to the and the users and then all the drones finish the delivery and then they will re return to

the truck and then the truck will return to the warehouse okay so this is the the status this the current

status uh as mentioned before for the current Logistics we are mainly using the trucks uh and the tricycles

trilers uh and for very small proportion we use the tracks plus the drums uh so take a look at this diagram

uh you know this is a a key study in Beijing uh you have the root from the warehouse to the distribution Point number one distribution Point number two

this is the track routs right number three number four uh and then go back to the warehouse so this is basically this is the roote of a truck route and then

you have the distribution Point connecting with the trilers and the trilers will uh finish the end users end

the user to finish the last La last segment uh delivery to the end users so you may ask why can't we use

drones to replace the trucks because we don't like the trucks the trucks are very slow to affect our social Vehicles affect the traffic flow the trailers are

very irregular vehicles and they affect the traffic flow substantially uh why why not using drones to replace the trucks and

trilers okay there are two reasons number one the regulations you know but we can try to convince the authorities uh but what are the tech technical challenge what are the tech

technical challenges uh there are two reasons number one a drone's driving range is

only around 10 me so which is inadequate for City range delivery the range is limited number two

the cost per kilometer per kilogram is higher than the truck filler combination to solve the problem to solve these two problem

uh my group we develop a new new approach a new device that allow the drones to land on the roof of vehicles

you see this diagram this is a drone the Drone can land on the rooftop of the vehicles the social Vehicles

while the vehicle are moving while all the vehicle are in motion with this tool we when we reach to could move a box from uh one area to

another area for example if if we separate the Beijing Beijing City into a number of zones you see Zone one is the

left uh left up corner and the Zone N is the right uh lower corner so you want to travel you want to uh move one puzle

from Zone one to don't end um here uh what we need to do is simply to find social vehicle you have numerous social

vehicle in one and you want to find one social vehicle that intends to travel from Zone one to Zone nend then we rent

the roof talk bases of this vehicle and the roof spaces are currently uned so we can pay a very small money to rent the

rooftop space of that St of that vehicle uh so for to to to finalize the long

distance delivery in between and we send a drone to to for the first small segment from the origin to the rooftop

of the vehicle and then uh the Zone will fly will take off to for other tasks and

the the the pule will stay in the Ro on the roof roof uh rooftop of the social vehicle and the social vehicle will take

their uh predetermined route and the original route from Zone one to Zone end as soon as the vehicle arrive at the Z

end you send another uh Zone to get the to to take the uh pass off and then for fly to the Final Destination right fly

for for the uh Final Destination that means for the first very small segment we use drones for the last very small segment we use drone for the very long

distance in between we use the currently wasted vehicle roofs so for you can imagine the problem is uh you know it's very cheap right

because the uh long distance in between you use the currently Ved vehicle rooftop uh roof vehicle roof that means

the you can pay a very small price very cheap price to to have this service now you saw we Sol the about mentioned two problem

number one we don't need to travel for long time because you only need to travel for the first very small segment

and the last very small segment not for 20 minutes even not for two or three minutes only for three for two very

small segment number two the cost is very cheap because the rental for the currently wasted car roof will be very

cheap uh this is the very uh very similar to the the car pooling system by The Uber

or by uh the DD in China the the equivalent of Uber in TR DD so very similar but uh the vehicle itself they

don't need to change the route right because for the car pooling you need to change the root pick up the pattern other pattern so the you need to change

the rout which is no good so but for this scenario this case Logistics of uh Uber system UB C ping system so you

don't need to change the rules because the the drones are flying faster than the vehicles and number two you don't need to share your cabin spaces with

another stranger and because the the pel on the rooftop uh you don't you don't even see it right you don't even see

it okay for each catchment because we we can separate the BP into a number of catchment and for each cat catchment we can allocate a few drones for uh the

first segment first service or for the last segment service for the end for the other catchment is the same and the the car the pass is carried by the vehicle

from catchment one to another catchment where the uh end user is located and another drone in end the user catchment

flies to the vehicle pick up the pel and deliver it to the end user so with this uh system if uh this remember this is the number one the

number one system is the current system in Beijing the number two system is the system uh in resarch papers in the

United States and number three is all system uh you know you can imagine you know all system will have the cost

benefit because we are using the currently visted rooftop spaces of social Vehicles so all cost is lower and

we will have the time because you far flying and you you solve The Last Mile first mile problem we have

a Time advantage and such more and more users will use this system instead of truck Traer combination uh in in my my

group we also have done the work using uh discrete Choice mode to estimate the market share for the uh current truck um

trailer combination and all system and then we can replace a lot of truck a lot of trailers which will improve the

traffic condition why because we didn't introduce any new vehicles right we didn't introduce any new vehicles and we

we replace many trucks no more trucks the the number of trucks are much much smaller than before number of trilers are not much smaller than before and we

are not introducing new vehicles because we are using the rooftop rooftop of the currently V State social vehicle

space so the traffic volume will be lower because you you have less number of trucks and you have less number of trailers we reduce the number of trucks

and trailers and we don't introduce new ground vehicles and in this in this way the last traffic on ground the the

traffic volume will be reduced uh then the highest speed the vehicles have will have right because less traffic then less crowded you have higher tra travel

speed then each and every passenger in the city will have their uh short time we save their time day by day that is a

huge benefit for the society that is a huge benefit for the society we also as estimate the social benefit for sa

annually okay uh the about example Okay I uh I will speed up because we we may not have enough time

for question answers so uh I uh the about example is for Logistics technological uh speaking there is no

challenge no much that challenge but for uh passenger transport uh in the vertical you know three dimension system

uh it is still we are still waiting but uh uh this example is about the passenger transport and we uh we

also have done studies for the passenger Transportation with flying vehicles actually it's not as flying vehicles is

a eval it's a electric vertical taking off and Light in devices um because the they are not able to drive they can only

fly they are not Land Air uh amphibious um here we we only consider e

um Z can only fly uh they cannot travel on road we design uh a Land Air cooperation system with the first it's

similar to the previous example with the first segment and the L segment using line transportation while the long

distance be in between using the uh EV the low altitude air transportation so you see for this catchment all the passenger you see the for this cat

catchment all the passengers will travel through landine transport from the hole to the water part that means the V point

the taking off space to the vti part and then you use the uh air transportation the low altitude air transportation from

One V point to another ver point and then you have the last segment you have the last segment by line transport okay uh to achieve this we

have to design the vertical right to determine where we allocate vertical vertial how to arrange the

traffic uh this is actually a very huge uh study that we Compares uh 12 largest city in China and 10 largest

metropolitan areas in the United States uh we also use the high resolution TR time data uh I'm not going to introduce

the technical details uh again I'm int I'm introducing a few findings a few findings okay I need to fet some water

withit me for 20 seconds okay I'm back so the first this is the vort uh

diagram you know you have quite a few vort in different cities in Chinese cities in the United States so uh Chinese cities are more

crowded right in Chinese City are more crowded than the cities in United States we thought at the very beginning we thought at the very beginning this St

system will work better for Chinese ciens and we move some trips up in the air right and then the traffic volume on the ground will be reduced and then we

will have short time uh you know more substantial reduction in China than the United States however at per

according to all study uh the cities in the United States states perform better in terms of TR Time Savings so roughly

around 40 to 50% depending on different metropolitan area but in China it's only 30 to 40% reduction in different

cities this is because this V part this vort you see this vti part they also induce some local travel demand uh which

makes our local Road more crowded there is no much difference between the two minutes flying time and the three minutes flying

time because even the distance is very very long but in travel time short uh so the total travel time will be dominated by

the first small segment by line transport and the last small segment by line land transport so the long distance

in in between is is short no matter it is three four five minutes uh flying time so the total Trav time for example

35 minutes will be dominated by the first segment and the last segment because you have this V part you have this V part uh so for the local traffic

it's more crowded and then the tri time for the first segment and the last segment in both China and United States uh are longer than before uh but the

total travel time have been red reduced for 30 to 40% in in China 40 to 50% in

theit St so uh as such what should we do what should we do we need to have the ground infrastructure adaptation to the

new system so we need to change the infrastructure nearby the what area otherwise uh the efficiency will not be

high uh okay another finding another Finding which is also very uh surprising uh because people always considered

battery the battery is the uh biggest challenge uh for the EV for the flying vehicles the battery is the biggest

challenge uh however according to all study even if the battery is more than adequate we don't set any constraints in

the battery uh we will only travel for three to six minutes in China and United States we don't need to

long time because the speed is high is within a city the distance is not that long so you don't need to travel too long time three to six minutes that

means if you have 10 minutes to 15 minutes and Battery side uh it will be more than adequate so we we don't need

to have long uh a large battery uh we don't need very long the city sze is very is relative small uh that is if we

we can develop batteries of 10 minutes range 12 minutes range that will be more than adequate even if we have 30 minutes

uh flying time for the urban area uh we will use small battery we don't need to use bigger battery because bigger battery that means heavy it's heavier no

good no good for the uh you know flying vehicles uh this diagram is uh the problem that I mentioned just now you see the green the

green uh links row that means those green links the traffic volume will be reduced and the red links that means the

traffic volume will be increased so is uh take a look at the demand representation uh and can be see clearly

in the diagram uh the long distance uh 12av uh is smaller you see all the long distance the it's green right it's

smaller but the local traffic for local room the volume have been increased the local room nearby the V vertical uh is

much biger than before so we need our road infrastructure as I mentioned now we need our road infrastructure adapted to the change before implementing the

new system okay to end uh today's uh three key studies uh is a very ly presentation so I need more

water uh I share with you a few thought a few a few of my thought um we are in the era of

uh uh re revolution in trans area we are in area of fast changing fast developing nowadays and we need

more Talent working in this area and you know we have three point even for the general

engineering the three points are very important number one uh you know engineering is increasingly interdisciplinary uh we need

interdisciplin collaboration between you know transportation engineering vehicle engineering electrical engineering computer science uh you know social

scientists even lawmakers because we are talking about the regulations of low altitude uh vertical transportation

system so in disciplinary is very important uh number two this is this is

a very practical discipline relying under uh government investment and also the government gets you know the toal we

have the huge social benefit that means the government gets the uh biggest uh benefit for example the uh you know if

we use drones to replace uh trucks and Killers is the the biggest beneficiary

is not the uh the Logistic Company uh the biggest the beneficiary is the government because uh all the citizens all the residents will enjoy shter time

time a shorter commuting time then the valuable time multiply their Time Time Savings that's a huge number of benef

social V okay number three is uh we need to look at the future development need uh we need to look at the future

development need we need to focus on the area of high growth we are we are not we we we shouldn't focus on what the

operation is now we should focus on what will the operation what what's the future operation will be in

future uh as I introduced earlier uh I taught uh in a few very different countries um you know and as the

professor uh and attended quite some outreaching outreaching activities with a young students with high school

students Senior High Junior High with uh junior Bachelor student undergraduate students so I found three points are

very important and number one you know happiness is actually generated from doing what you allow to do not by the return When you earn a lot of money

right happiness the Gen happiness is actually generated from doing what you all to do uh we need to find uh our gen

interest and follow your genuine interest uh and also the uh Perfection you know in Chinese we call

it aiming at the absolute perfection so this is another very important deciding factor for your success uh often use this diagram you see this diagram you

see this diagram I have often used this diagram in the in my lecture to my PD students uh this uh black curve this

black curve is the our current boundary of knowledge uh within the boundary we know the knowledge outside of the

boundary is unknown knowledge it know knowledge so if you are at the boundary if you add the knowledge boundary on

some opportunities on some occasions we may find a way to make break RS however if we are not aiming at the best we are

close to the boundary if you are here not close the boundary but not at the boundary not on the boundary uh even if you have opportunity

you have you have that opportunity there is no le for you to make a break through so and we are not at the boundary yet so your

perfect product is your call competitiveness uh number three is diversity is this is the reason why I

allow to uh attend the event like this uh because um we need to meet more people we need to meet more people we

need to talk more people uh especially to meet people from different countries different background uh different culture uh because we want to have

diversity diversity generates creativity uh dver at a homo homogeneous environment your mindset is very likely

to be uh confined to be fixed without sying out of the box uh abilities okay this is the three that I want to give to

you so uh that's all for my presentation uh I'm happy to take any questions if you have um so if you you

can unmute yourself you ask a question or you can raise your hand use a re hand function of of Doom and then the moderator will call your name you ask

questions uh alternatively you can also uh you know typ key in type in your your question in the chat

box okay if you have any questions let me know