Patent Application
20240210962
The invention relates to an autonomous transportation network and method for operating the same.
The term “automated transit network” or “automated transportation network” (abbreviated to ATN) is a relatively new designation for a specific transit mode that falls under the larger umbrella term of “automated guideway transits” (AGT). Before 2010, the name “personal rapid transit (PRT)” was used to refer to the ATN concept. In Europe, the ATN has been referred to in the past as “podcars”. This document sets out the ATN concept and describes a novel method of operation of the ATN concept.
Like all forms of AGT, ATN is composed of automated vehicles that run on an infrastructure and are capable of carrying passengers from an origin to a destination. The automated vehicles are able to travel from the origin to the destination without any intermediate stops or transfers, such as are known on conventional transportation systems like buses, trams (streetcars) or trains. The ATN service is typically non-scheduled, like a taxi, and travelers are able to choose whether to travel alone in the vehicle or share the vehicle with companions.
The ATN concept is different from self-driving cars which are starting to be seen on public streets. The ATN concept has most often been conceived as a public transit mode similar to a train or bus rather than as an individually used consumer product such, as a car. Current design concepts of the ATN currently rely primarily on a central control management for controlling individually the operation of the autonomous vehicles on the ATN.
On the other hand, the self-driving cars are often described as being “autonomous”, but in practice, there are different classes or levels of vehicle autonomy. The degree of vehicle autonomy is typically divided in five levels, as set out by the On-Road Automated Driving (ORAD) committee of the Society of Automotive Engineers (SAE) in “Taxonomy and definitions for terms related to driving automation systems for on-road motor vehicles” published in Recommended Practice SAE J 3016 on 15 Jun. 2018. Level 0 refers to a vehicle that has no driving automation. The driver of the vehicle is fully in charge of operating the movement of the vehicle. Vehicles of Level 0 may include safety systems such as, for example, a collision avoidance alert. Level 1 refers to vehicles having at least one driving assistance feature such as an acceleration or braking assist system. The driver is responsible for the driving tasks but is supported by the driving assist system which is capable of affecting the movement of the vehicle. Level 2 describes vehicles having more than one assist system for actively affecting the movement of the vehicle. The driver, in Level 2, is still responsible for the driving tasks and must actively monitor the trajectory of the vehicle at all times. The driver is, however, actively supported by the assist systems. Level 3 describes a so-called “conditional automation” of the vehicle. The vehicle is capable of autonomously driving in certain situations and with limitations. The driver is not required to actively monitor the assist system but is, however, required to take control of a driving situation if requested by the assist system. Level 4 describes autonomously travelling vehicles which are capable of travelling specific routes under normal conditions without human supervision. The vehicles of Level 4 can therefore operate without a driver but might need remote human supervision in case of conflict situations, travelling in remote areas, or when travelling extreme weather conditions. Level 5 Automation describes fully autonomously driving vehicles. No human interaction is required at any time for the operation of the vehicles.
The reliance of the existing ATN networks on a central control management leads to a bottleneck in that each of the autonomous vehicles needs to be in almost continuous communication with the central control management. This can result in problems if the communications network is overloaded or there is a major incident somewhere in the ATN network that requires action from the central control management.
An example of such as central control management is outlined in U.S. Pat. No. 10,345,805 (Seally, assigned to Podway Inc.) in which the central control management receives a request from an autonomous vehicle for a route from the origin to the destination. The central control management calculates the route and sends to the autonomous vehicle a journey instruction set to allow the autonomous vehicle to navigate from the origin to the desired destination along the calculated route. The central control management in this system needs to transmit large amounts of data from the autonomous vehicles and gather data from the autonomous vehicles on a continuous basis. This requires a large amount of hardware and data bandwidth and can cause a problem if an autonomous vehicle enters an area in which connectivity is poor. In the event of a breakdown of the central control management, then the autonomous vehicles will no longer be able to navigate or recalculate journeys.
Many current ATN concepts rely on guideways being built as part of the infrastructure. This may have its advantages when dedicated infrastructure separate from other traffic flows or pedestrians can be designed. The cost of the provision of the guideways is significant and this will delay the development of the ATN network. One example of such a guideway is the infrastructure that can be seen in London Heathrow airport's Terminal 5.
A report on “Automated Transit Networks (ATN): A Review of the State of the Industry and Prospects for the Future” published by the Mineta Transportation Institute, Report No 12-31 in September 2014 reported that at the date of writing no ATN having more than ten stations had been implemented in the world. Currently the ATN networks operate on the principle of mapping each origin to all of the destinations. This leads to a matrix with 20 entries even for a simple five-station system as there are four possible destinations from each of the five origins. A ten-station system would have 90 possible routes and it will be seen that as the number of origins and destinations increases, then an O/D matrix listing all of the possible routes will expand out of hand. The current systems are therefore not scalable.
A further issue that has been identified in the ATN network is the handling of multiple vehicles and prioritizing of access for priority vehicles, such as paramedics or police. A solution is offered in U.S. Pat. No. 9,536,427 (Tonguz et al, assigned to Carnegie Mellon). The solution uses vehicle-to-vehicle communication to establish a priority zone as required.
U.S. Pat. No. 10,152,053 B1 discloses an autonomous vehicle management system and a method for the controlling of a fleet of vehicles. The system comprises a plurality of autonomous vehicles having an onboard processor and vehicle memory for calculating of a route. The system further comprises a control management center and a plurality of beacons for communication between the control management center and the autonomous vehicles. The beacons may also be used for determining a location of the autonomous vehicle within a transportation network. The method comprises receiving a plurality of requests for rides in the autonomous vehicle by passengers, identifying demographic information related to each of the plurality of passengers, determining a vulnerability score and a priority for each of the plurality of the passengers; and causing a particular one of the autonomous vehicles to pick up the passengers. The vulnerability score is used to assess the vulnerability of the passenger in assigning and to calculate a priority to pick up each passenger. For example, passengers in higher crime areas and/or having demographics matching a vulnerable subset (e.g., senior citizens) of the population may be deemed to be more vulnerable and may be assigned higher vulnerability scores. The U.S. patent further discloses sending and receiving of traffic information by the autonomous vehicles using the beacons. An independent calculation of the routes in the control management center and the onboard processor is not, however, disclosed.
US 2009/037086 A1 describes a method for equalizing traffic flows for autonomous vehicles in a transportation network using a control computer. The control computer stores information on the routes in the network as vectorized graphs representing the routes extending from an origin to a destination. The route network contains multiple branch points creating separate branches of the routes for the travelling of the autonomous vehicles. At least some of the autonomous vehicles send a unique vehicle identifier and their current position to the traffic control computer. The information sent by the autonomous vehicle includes the destination(s) to which the autonomous vehicles are travelling. The document also discloses a method for producing and transmitting a route recommendation to the vehicles. The route recommendation comprises sending a distribution ratio V to the vehicles. The autonomous vehicles themselves then calculate the alternate routes from the origin to the destination using a randomized selection scheme using the distribution ratio V. The document also teaches a method for transmitting the individual route recommendations to the vehicles. This method comprises briefly connecting a communication system, mounted close to the road via wireless communication to the autonomous vehicle. Different route recommendations are alternately sent to the passing autonomous vehicles. The U.S. patent application remains silent on the issue of an independent route calculation by the control management center and an onboard processor of the autonomous vehicles.
There is therefore a need for providing a resilient autonomous transportation network.
An autonomous transportation network is disclosed. The autonomous transportation network is aware of the movement of all autonomous vehicles travelling within the autonomous transportation network without the autonomous vehicles having to constantly communicate the position in the network to a control management center.
The autonomous transportation network comprises a plurality of autonomous vehicles with an onboard processor and vehicle memory for calculating a route between an origin and a destination and a vehicle antenna for transmitting the calculated route. A control management center for control of the transportation network comprises a control management processor and a central memory. A passenger can request to travel the route from an origin to a destination in the transportation network. The control management processor independently calculates the routes of the plurality of autonomous vehicles from the origin to the destination. A plurality of beacons, located for example at junction, is connected to the control management center and receives redirection information from the control management center for transmission to one or more of the plurality of autonomous vehicles.
This network enables the plurality of autonomous vehicles to travel along the route from the origin to the destination independently. The control management center is aware of the movement of the plurality of autonomous vehicles travelling in the autonomous transportation network without the autonomous vehicles having to constantly communicate with the control management center. This independent calculation of the routes hereby reduces the communication overhead of a communication infrastructure.
The control management center is adapted to determine a conflict situation on the routes of the plurality of autonomous vehicles.
The control management center is further used for simulating a traffic demand and the routes of the autonomous vehicles and for determining corrected route instructions for the autonomous vehicles in case of a necessary redirection along the route.
A method of operation of an autonomous transportation network comprising a plurality of autonomous vehicles is also disclosed. The method comprises receiving an instruction for a journey from an origin to a destination, calculating in at least one of plurality of autonomous vehicles a route from the origin to the destination, independently calculating in a control management center the route from the origin to the destination, comparing the route calculated in the one of the plurality of autonomous vehicles with the route independently calculated in the control management center and, in the event of a disturbance, sending corrected route instructions to the one of the plurality of autonomous vehicles. The sending of the corrected route instruction comprises sending the corrected route instructions to one of more beacons. The corrected route instructions can be one or more of speed instructions or diversion instructions.
A method for calculation of a route from an origin to a destination in an onboard processor of an autonomous vehicle comprising receiving an instruction for a journey from a control management center. The received instruction comprises the origin and the destination of the journey and is relayed by, for example, the beacons. The method further comprises locally calculating a direct route from the origin to the destination. Locally calculating the route is done using the onboard processor and a geographic data stored in a vehicle memory. In a further step, the method comprises receiving a corrected route instruction by an antenna of the autonomous vehicle from a control management center. The corrected route instruction is addressed at the one of the plurality of autonomous vehicles. The onboard processor of the autonomous vehicle locally recalculates a new best route for a remainder of the route to the destination after receiving of the corrected route instructions. Calculating the new best route is done using the onboard processor, the geographic data stored in the vehicle memory, and the received corrected route instructions. The autonomous vehicle subsequently continues along the corrected new best route. The autonomous vehicles include assist systems of Level 2 or Level 3, as described above. Using the vehicle geographic data and the onboard processor, the autonomous vehicles are capable of autonomously travelling in the transportation network. The autonomous vehicles, therefore, do not require a driver for driving of the autonomous vehicle.
A method for calculation of a route from an origin to a destination in a control management processor comprises receiving an instruction for a journey. The instructions for the journey comprise the origin and the destination of the journey. The control management center assigns one autonomous vehicle of the plurality of autonomous vehicles for fulfilling the request of the passenger. The control management processor independently calculates, using the geographic data stored in the central memory, the route to the destination.
The autonomous vehicles 20 can be parked in a parking place with a plurality of tracks 15 or be in motion along the tracks 15. The autonomous vehicles 20 will be typically battery powered and can be charged, for example, when they are in the parking places.
The autonomous transportation network 10 has a control management center 100 which monitors the progress of the autonomous vehicles 20 but does not directly control the progress of the autonomous vehicles 20, as will be explained below. The autonomous vehicles 20 can send and receive information to the control management center 100, if necessary, and are connected to the control management center 100 through wireless connections using a vehicle antenna 25 located on the autonomous vehicle 20 in communication with the control management center 100 through the communications antenna 110 at the control management center 100. The control management center 100 is provided with a processor 120 and a central memory 140. The control management center 100 is connected to the beacons 17 using fixed communication lines 105 (although of course it would be possible to also use wireless connections over the distance between the beacons 17 and the control management center 100 or over part of the distance if required). The central memory 140 includes central geographic data 124 about the autonomous transportation network 10 including the location of the beacons 17.
The autonomous transportation network 10 is provided with a plurality of stopping points (also termed stations), as is known from a railway, tram, or bus network. The stopping points will be clearly labelled to passengers 35 who wish to use the autonomous transportation network 10. A vehicle memory 28 in the autonomous vehicle 20 stores a vehicle geographic data 24 in the form of a network map with the location of the plurality of stopping points and also a selection of precalculated routes along the tracks 15 between any two of the stopping points. There will generally be more than one pre-calculated route between two of the stopping points to allow for alternative paths to be followed, as will be explained later.
The autonomous vehicle 20 has not only the afore-mentioned vehicle antenna 28 and the vehicle memory 28 but will also include an onboard processor 27 which can control the autonomous vehicle 20 using the information in the vehicle memory 28 and any information received from the beacons 17.
Suppose now that a passenger 35 at a first one of the stopping points, termed an origin 30, wishes to travel to a second one of the stopping points, termed a destination 40.
The request 37 is received in step 220 by the control management center 100. The request 37 will include details about the origin 30 of the passenger and the planned destination 40 of the passenger. The origin 30 can be determined by either using GPS coordinates transmitted in the request 37 from a smartphone or by transmitting the number of the stopping point in the app. The destination 40 of the passenger 35 will be determined in step 225 by either inputting the number of the stopping point corresponding to the destination 40, or an address of the destination 40 or selecting a point representing the nearest stopping point to an address on a map displayed on the screen of the smartphone.
The control management center 100 stores the data received through the request 37 concerning the origin 30, at which point the passenger 35 wishes to be picked up, and the destination 40. The control management center 100 will then generally assign in step 230 the autonomous vehicle 20 closest to the passenger 35 to pick up the passenger 35 from the origin 30. It will be appreciated, of course, that there may already be one of the autonomous vehicles 20 at the origin 30 and the passenger 35 may in fact be standing next to one of the autonomous vehicles 20 and other ways of communication, such as NFC communication or by scanning a bar code or QR code on the vehicle could be used to reserve the autonomous vehicle 20 for use by the passenger 35. These examples are not limiting of the invention.
The autonomous vehicle 20 will then calculate in step 240 locally in a local processor 27 using the vehicle geographic data 24 (network map plus precalculated routes between the stopping points) stored in the vehicle memory 28 the route 50 to the destination 40 to which the passenger 35 wishes to go.
At around the same time in step 250 the control management system 10 will independently calculate using the control management processor 120 the route to the destination 40. The vehicle geographic data 24 stored in the autonomous vehicle 20 is identical or substantially similar to the central geographic data 124 stored in the central memory 140 and thus the control management system 10 will know the route that the autonomous vehicle 20 will take between the origin 30 and the destination 40. In other words, the central geographic data 124 stored in the central memory 140 comprises identical or similar data compared to the central geographic data 124 stored in the autonomous vehicle 20. The central geographic data 124 might comprise, however, more detailed data as, for example, the simulated current traffic situation in the transportation network 10. Hence, the route calculation in the autonomous vehicle 20 and the route calculation in the control management system 10 will be performed separately from each other in real-time based on the vehicle geographic data 24 and the central geographic data 124 and will initially not take into account any disturbances, such as but not limited to traffic accidents, traffic jams.
Once the route 50 has been calculated in the local processor 27, the autonomous vehicle 20 will start its journey from the origin 30 to the destination 40. Unlike in prior art systems, the autonomous vehicle 20 is not required to notify the calculated route 50 to the control management center 100. The control management center 100 knows, as described above, the route of the autonomous vehicle 20 by calculating the route 50 in step 250.
The purpose of this dual calculation of the routes is to enable the control management center 100 to determine what is happening in real-time in the autonomous transportation network 10. There will not be a single passenger 35 requesting a single one of the autonomous vehicles 20, but a number of passengers 35 requesting a number of autonomous vehicles 20 from a plurality of the origins 30 and going to a plurality of the destinations 40. It is the role of the control management center 100 in step 260 to simulate the traffic demand and the routing of the autonomous vehicles 20 and, if necessary, make changes of the routes 50 or adjust the speed of travel of the autonomous vehicle 20 as will be described in more detail in the examples set out below.
In the event that the control management center 100 determines that the autonomous vehicle 20 needs to deviate or needs to be redirected from the calculated route 40, then the control management center 100 sends corrected route instructions 50cor. The control management center 100 does not send these corrected route instructions 50cor directly to the autonomous vehicle 20, but in step 270 corrected routing information is sent to one or more of the beacons 17 which can then redirect or slow the autonomous vehicle 20 in step 275.
The communication between the beacons 17 and the autonomous vehicles 20 is carried out locally and does not require much power. Only those beacons 17 near the position of the autonomous vehicle 20 need to be provided with corrected routing instructions 50cor to be received by individual ones of the autonomous vehicles 20. Unlike in prior art systems, only individual ones of the autonomous vehicles 20 need to change the route 40 if a possible conflict is detected. The control management center 100 knows, from independently calculating in step S250, the location of the autonomous vehicle 20 in the autonomous transportation network 10. The control management center 100 therefore only needs to inform those beacons 17 near the position of the autonomous vehicle 20 of the corrected routing instructions 50cor. The local transmission of information between the beacon 17 and the autonomous vehicle 20 also reduces the risks of hacking of the autonomous transportation network 10 as the amount of data transmitted is very small and the distances of wireless transmission are also short.
These corrected route instructions 50cor will ensure that the autonomous vehicle 20 changes the route 50 or to alter its speed, as will be explained below. The autonomous vehicle 20 after redirection will recalculate (as in step 240) the new best route 50new to the destination 40 using the vehicle geographic data 24 and continue the journey along the corrected new best route 50new to reach the destination 40. The control management center 100 will also be able to determine the new best route 50new and will then be able to simulate the route (step 260) to determine whether there are further issues that may need a further redirection of the autonomous vehicle 20.
An example of a necessary correction to the originally calculated route 50 is shown in
There is no need for the control management center 100 to broadcast to all of the autonomous vehicles 20 in the autonomous transportation network 100 information about the blocked route at the position 55. Only those autonomous vehicles 20 that have calculated the direct route 50dir which passes through the blocked position 55 will receive the redirection information locally from the beacon 17. This eliminates much of the potential data traffic sent from the control management center 100.
The vehicle memory 28 in the autonomous vehicle 20 does not need to store unnecessary information about the blocked routes. This simplifies the calculation of the new route 50new in the onboard processor 27 which results in a quicker calculation with the use of fewer resources. The vehicle memory 28 can be kept smaller.
The amount of resources used the control management center is also reduced since the control management processor 120 only needs to inform the beacons 17 at the start junction 56 of the blocked route that there is an obstruction due to a broken-down autonomous vehicle 20′. There is no need to broadcast the information to all of the autonomous vehicles 20.
A further example of the efficient management of the autonomous vehicles 20 is shown in
The control management center 100 is able to send information to the autonomous vehicles in step 270 to the autonomous vehicles 20a-d using the beacons 17x and 17y located near the entries to the roundabout 57. The information will not be needed to travel along another route 50alt, as shown in
The autonomous vehicle. 20 recalculates, in step 330, the new best route 50new for the remainder of the route 50 of the autonomous vehicle 20 travelling to the destination 40. The autonomous vehicles 20 include assist systems of Level 2 or Level 3, as described above. The autonomous vehicle 20 is capable of autonomously travelling in the transportation network, using the vehicle geographic data 24 and the onboard processor 27. The autonomous vehicles 20, therefore, do not require a driver for driving of the autonomous vehicle 20. The recalculating of the new best route 50new is done by the onboard processor 27 using the vehicle geographic data 24 stored in the vehicle memory 28 and the received corrected route instructions 50cor. The autonomous vehicle 20, in step 340, continues the journey along the corrected new best route 50new to the destination 40.
The control management processor 120 of the control management center 100 independently calculates, in step 520, the route 50 of the assigned one of the plurality of the autonomous vehicles 20 from the origin 30 to the destination 40. Calculating the route 50 by the control management processor 120 is done using the central geographic data 124 stored in the central memory 140. In step 530, the control management center 100 simulates a traffic demand and the routing of the autonomous vehicles 20 in the transportation network 10, using the requests received from the plurality of passengers 35 and the control management processor 120. The control management center 100 determines, in step 540, the corrected route instructions 50cor to enable the redirection of the assigned one of the plurality of autonomous vehicles 20 from the route 50. Determining the corrected route instructions 50cor is done using the simulated traffic demand.
The corrected route instructions 50cor are, in step 550, relayed to the assigned one of the autonomous vehicles 20 for redirecting the autonomous vehicle 20 to the alternative route 50alt or for recalculating a new best route 50new by the autonomous vehicle 20, as elaborated in the description of
| Number | Date | Country | Kind |
|---|---|---|---|
| 101673 | Feb 2020 | LU | national |
| 2003395.7 | Mar 2020 | GB | national |
This application is a US national phase entry of international patent application No. PCT/EP2021/052377 filed on 2 Feb. 2021 and claims priority of Luxemburg Patent Application number LU101673, filed on 9 Mar. 2020 and UK Patent Application 2003395.7, filed on 9 Mar. 2020. The entire disclosures of the Luxemburg Patent Application LU101673, the UK Patent Application 2003395.7 and International Patent Application No PCT/EP2021/052377 are hereby incorporated herein by reference.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2021/052377 | 2/2/2021 | WO |
