Background to the invention

This invention relates to a method and apparatus for intelligent and connected in- vehicle navigation to one or more destination bays for the purpose of parking, vehicle charging or other waiting event. The navigation directions are updatable responsive to information sources which may comprise bay location, bay occupancy, approach direction or traffic flow circumstances, including but not limited to dither time of other parking vehicles.

It is well known that parking in large car parks can give rise to queues. For significant events such as concerts, exhibitions or sports events this can cause traffic queues affecting a wide area both in or near the carpark and in the surrounding road network.

It is known for car parks to monitor entry and exit and display available space on signs close to or within the car park or floor occupancy in a multi-story car park, and drivers can take account of that information when deciding where to try and park. However, for large events such as the British Grand Prix at Silverstone, traffic is disrupted for several kilometres around. Some parking systems such as those for long term parking at airports or ports may use zoning directions. There are also various everyday locations such as work car parks with 500, 1000 or more vehicles arriving at similar times and congestion being caused as parking queues form while bays are sought and maneuvered into. Similar congestion may also happen at other vehicle parks, for example in freight or lorry parks. Large shopping areas also experience queueing vehicles especially at peak times.

Queueing is environmentally unfriendly and at present costs, half an hour in a queue can use more than a litre of petrol. Increasing use of electric vehicles will also create demand for charging while parked, requiring new strategies to identify and organize charging availability. Some estimates suggest at least 10% of cars will seek to charge when parking. Behavioral preferences by c rivers such as parking as close as possible to an entrance or exit or some other choice or pre-allocation (including EV charging points) create patterns that lead to queues especially when many vehicles are massing for a time sensitive event. Drivers find themselves waiting while other vehicles in front select, manoeuvre and park. The ‘dither tine’ and behavioural patterns both contribute to queue formation and queuing time.

The present invention is directed towards an intelligent and connected parking system that reduces the average wait time of vehicles by coordinating where and how vehicles park in the vehicle park.

It is known frcm the document US2020/0C 18602 A1 to orovide a system for the control of parking of vehicles, and for use with vehicles which include or contain apparatus for receiving directional information for the guidance of a vehicle and for providing signals denoting the current location of the respective vehicle. The system comprises a database which includes a representation of a vehicle park, the representation including parking lane entrances, parking lanes extending from the lane entrances and parking bays along the lanes; and a computerised controller which for each participating vehicle provides respective directional information. The controller is disposed to distribute incoming vehicles among the parking lanes and for any lane to direct a succession of vehicles along the lane to respective bays or groups of bays so that a majority of the vehicles are each directed to a bay or group of bays that is spaced apart along the lane from the bay or group of bays to wiich the preceding vehicle has been directed. The purpose of that parking scheme is to allow sufficient pathways between vehicles across the parking lot.

The present invention is directed towards an intelligent and connected parking system that reduces the average wait time of vehicles by coordinating where and how vehicles park in the vehicle park.

The present invention aims to allow successive vehicles space to manoeuvre individually without interfering with one another.

In one form of the invention the the controller is disposed to direct a set of vehicles along a lane such that the vehicles of the set are sequentially directed to bays or groups of bays progressively close' tc the entrance by which the vehicles enter that lane.

The invention also provides a system for the control of parking of vehicles, and for use with vehicles which include or contain apparatus for receiving and display ng directional information for the guidance of a driver and for providing signals denoting the current location of the respective vehicle, in which system! vehicles are tracked and after entering the proximity of a vehicle park they are given instructions directing them to a specific parking lane and parking bay or set of bays in a sequence that for every lane allows successive vehicles space to manoeuvre individually without interfering with one another, whereby to disrupt known driver preference patterns.

The invention also provides correspon ing methods.

In preferred embodiments of the invention vehicle locations are monitored via an onboard communicating device, which may include or be used in conjunction with satellite navigation, or by other positioning detection such as Bluetooth or mobile phone positioning software and directed in real time (or near real time) to a set of bays in which they can park with minimal delay taking into account both occupancy of the car or vehicle park and parking movements of other vehicles. A processing unit analyses the information from vehicles and/or parking bays indicating where vehicles have been parked, where those that are in the stages of parking are manoeuvring and location of vehicles that are intending to park in order to determine the parking strategy for individual vehicles. The vehicle-specific strategy is then displayed in the vehicle for use by the occupants.

This system disrupts driver behaviour patterns that lead to vehicle clustering and by providing specific instructions to follow, dither and decision time may also be reduced or el minated. The intelligent system may also monitor vehicle behaviour, any emerging behaviour patterns anc update the modelling on which instructions are based.

The invention is defined by the claims. Brief description of the drawings

The invention is now described by way of example with reference to the accompanying drawings in which:

Figure 1 is a schematic illustration of the formation of a queue of vehicles in a vehicle park when vehicles look to park near the entrance to a venue;

Figure 2 schematically illustrates row parking allocation according to intelligent parking information in order to break queues and avoid tail events;

Figure 3 schematically illustrates spaced apart bay parking allocation according to intelligent information flow and reduction of queue length;

Figure 4 further schematically illustrates vehicle parking according to intelligent information flow to avoid tail events;

Figure 5 is a flow diagram of information and processing events according to an embodiment of the invention;

Figures 6 and 7 schematically illustrate queue ticket triggering when in vehicle park proximity;

Figure 8 is a schematic illustration of an in-vehicle display;

Figures 9A and 9B show virtual signing display instructions integrated with satellite navigation;

Figure 10 shows a schematic illustration of a node and line segment diagram representing stored vehicle park data;

Figure 11 shows a more complex branched node and line segment diagram;

Figure 12 illustrates various parking bay types, and

Figure 13 illustrates a node and segment diagram corresponding to the vehicle park layout shown schematically in Figures 1 to 4 Detailed description

Queues form in vehicle parks for a variety of reasons and a predominant reason stems from waiting while other vehicles perform a parking action. This can happen even when the vehicle park is far from full and is very disruptive for major events or when a lot of parking happens at the same time. The formation of a queue of vehicles is called herein a ‘“ail Event’.

In the parking system of a preferred embodiment of the present invention dynamic vehicle position monitoring, in combination with stored information that includes destination and vehicle park spatial layout, is utilised to provide vehicle-specifc instructions and direction to a specific parking bay or set of bays in a manner that minimises delays and interference due io other vehicles as they manoeuvre to park.

Vehicles participating in the system connect via an on-board application that can run on a communicable device such as a phone or in-built system and either includes or links to a navigation system, which may be satellite or other geo-positioning system. Other information that may be utilised includes vehicle type and required parking bay type, for example ‘disabled", ‘EV charging’ or ‘young family’, and derived information concerning vehicle behaviour patterns including queue formation, accident information or other delay or dithering due to manoeuvres that may or have created a Tail Event.

The system may also utilise other available traffic information for navigation purposes prior to entry into the vehicle Dark and parking phase. Preferably the monitoring of participating vehicles commences significantly in advance of entry to the vehicle park and is corrtiined in end-to-end navigation although systems which start more proximate the vehicle park entry are also envisaged. The system may also be augmented by Bluetooth or o:her sensory signaling within vehicle parks.

In the following description of embodiments of the invention the vehicles and park are described based on cars, but the system may also be adapted for use with other vehicles and massed parking, for example lorry parks, and may have different or additional features dependkig on vehicle type. In general, it can be assumed that vehicles sequentially and randomly arrive at a car park and tend to follow one of behavioural patterns:

• Park nearest venue entrance (Best Position)

• Park nearest car park exit (Best Exit)

® Park in first space available (First Open)

• Park randomly (Random)

If drivers and their vehicles follow the Best Position and/or Best Exit and/or First Open patterns the average parking time, taken as the mean average between arriving at the car park and turning the engine off to park, is longer than could be achieved if parking were coordinated to avoid obstructions and interactions due to manoeuvring. It is also possible that over time other behaviour patterns may emerge, for example due to the requirements of electric or driverless vehicles or the geometry of a particular vehicle park.

A simple embodiment of the invention is explained with reference to Figures 1 to 7 where it is assumed cars are seeking to park in Best Position nearest to the venue entrance. The principles elaborated can be adapted to take into consideration vehicles following other and multiple behaviour patterns.

In Figure 1 a plurality of cars 1 are all seeking to park near an Event Entrance 2 and a queue has formed shown on the right-hand side and along the bottom of the figure. The queue could extend beyond the carpark and into the nearby road system. This scheme could also illustrate the situation that arises in a section of a car park close to the exit where vehicles form a Tail Event attempting to park in Best Exit positions.

Figure 2 illustrates a strategy to unblock or avoid a Tail Event. The existence of the Tail Event could be detected from the monitoring the movement of participating cars (and may include assumptions therefrom about other cars) or by sensors in the carpark when available. A Tail Event will be associated with insufficient movement of cars.

The system having detected a Tail Event as in Figure 1 , various cars, via their drivers or other occupants carrying a connecting device, are directed to park in specific locations via their in-vehicle apparatus which informs them to turn down an earlier parking row. In the case of driverless vehicles, the information may be provided to the driverless system and may not need actual visible display. The directions to the in-vehicle apparatus come from the controller or controllers in the system, which is monitoring and coordinating vehicles, following what happens in a dynamic way, usually in real time.

This is illustrated in Figure 2 where cars V1 and V3 are directed to go to Row B and cars V2 and V4 are directed to go to Row C. Other cars will be similarly directed either as a continuing removal of the Ta i Event or as a precautionary measure to prevent or minimise Tail Events but still allowing some vehicles to flow towards the event entrance unless it is full.

Figure 3 shows the situation once cars have been diverted into rows as in Figure 2. Having entered a row, the controller directs each car to a specific bay or set of bays. In the drawing car V2 in Row C is sent to bay S2 deeper in the row and car V4 to bay S4 shown here in the middle of the row. This sequence means that the cars sent into the row do not have to pass preced ng cars parking in the row. It will be noted the bays allocated to the parking cars are not adjacent, indeed they are deliberately spaced apart to give the cars space to manoeuvre individually without interfering with one another as might happen if adjacent or opposite bays were assigned.

The cars V1 and V3 are similarly sent sequentially into spaced apart bays S1I and S3 in row B, each car being sent further in from the row entrance than the car behind it. Given the spacing apart of the cars the next set of cars into a given row will generally not be impeded as the preceding cars will also have had time to park. Control of the row directions may use various techniques including an expected time to park basis or monitoring of location, or a combination of both. As vehicle type is known it is also possible to allow extra time for example for large or long vehicles that take longer to manoeuvre.

It will be appreciated that this arrangement can scale up with many more row's, longer rows and potentially more vehicles sent into the row in the sets. Rows may also have bends and the term row or parking lane is used to indicate the path on which parking bays are located. The pa'ameters of the park are known and stored, and the algorithm can determine the number and frequency of turns into a given row and distribute that as individual messages to the participating vehicles.

StiH referring tc Figure 3, the early turn off creates headroom hi in the queue ahead of the next car V5 which is now no longer waiting for cars in front to complete manoeuvres. Row A will continue to fill either by cars not following instructions (there will always be some) ¹ or by way of instruction, or instructions can cease and allow free choice, for example in lighter traffic flows.

Although the invention may be used to alleviate Tail Events the preferred operation mode is anticipatory rather than reactive in order to prevent or reduce the occurrence of Tail Events. This is especially preferred for mass events where traffic flow will be high. Vehicles with requirements for special bays can also be given individual direction to their special bays from the moment that they approach the carpark.

The anticipatory mode is shown in Figure 4. Sets of vehicles are diverted into their allocated rows in the ‘first in furhest bay’ allocation sequence previously described. A time interval where no vehicles go into a given row may be created as subsequent vehicles are diverted to other rows to ensure the last in car of a set is parked, or the flow may be more joined up to a smooth distribution, depending on the various parameters concerned. This flow control allows multiple parking manoeuvres to take place within separated rows an: separated within rows with the vehicle set distribution sequences allowing row manoeuvres to clear, or substantially clear, before more vehicles are sent in. In the drawing vehicle V6 will be directed forwards until it reaches where it is allocated to a row. In the Figure the vehicles on the approach are shown more spaced apart than in Figure 2 to illustrate that a Tail Event has been prevented or at leas: reduced.

Overall, as well as removing, preventing or reducing entry queue Tail Events, the allocation and direction of veh cles to spaced apart parking bays within a row enables both less interference or waiting for other vehicles’ parking manoeuvres and reduction of searching dither tirre because there is no need for a driver to continually look out for vacant bays, cruise around as the car park becomes fuller or even choose between bays. Actual parking data is fed back to the system to take account of mis-parking, spontaneous actions by the drivers or by actions of drivers that are not in participating vehicles. The algorithm directing vehicles constantly updates in accordance with the occupancy and distribution state of the car park as well as vehicle flow. Other assumption modelling may also be updated, including by Al or other learning mechanisms, either dynamically or in periodic updates.

Various assumptions can be made from information to allow for rogue or nonparticipating drivers and traffic data may also be used to aid in those assumptions.

When the car park is not close to maximum occupancy the bay allocation may include giving a driver choice of more than one bay - for example two or three adjacent, opposite or otherwise close bays. Thus, if a ‘rogue’ vehicle has parked in one of the bays there is still a bay available at the relevant location. Indications may be given to the driver that there is choice in the set of bays, or there may be sequential directions to the first, then second and any subsequent spaces so as to avoid or minimise introducing decision time dither.

The system may also record when an offered space is rejected and treat that space as occupied and cease to make further allocations. Usually that will be done after a space is rejected a plurality of times. The spacing between allocated bays is preferably 4 or 5 spaces although this may depend on the layout of the car park. Where a set of bays is allocated to a vehicle, such as two or three adjacent bays, then the spacing between sets of bays may be reduced but there will still be some spacing. The spacing between allocated bays does not have to be empty bays and may contain parked vehicles, which will be the case as the park fills up.

For optimum effect the directions start before arrival at the park. This is particularly important for large events where tailbacks may cause significant traffic disruption and there may be multiple parks, more complex layouts and multiple access points. The initial navigation towards the par* may be any navigation system or an integrated one. The navigation used -or parking is more granular than many road navigation systems and at some point granularity will be switched (in an integrated system) or be displayed separately.

Figure 5 shows a schematic diagram of information flow in an embodiment of the invention. Users of the system will ge nerally have entered their parking intentions well before arrival and the pricing model for parking can include incentives to park in accordance with instructions and nence create a predominance of participating vehicles. This may all form part of the information stored in the system.

Once the in-vehicle application starts to run, for example the start of the journey or at a set time, it monitors for an information trigger and commences corresponding operations when a trigger is detected.

There may be several geo-position related information triggers and corresponding operations and displays. A first ’.rigger event may occur when the vehicle approaches the proximity of the park and crosses a location-based system (“Geofence”) or time- to-arrival algorithm, which may extend over a wide radius several kilometers from the venue. At this point a ‘Queue Ticket’ may be assigned to the vehicle as represented by box 51 in Figure 5.

Following the first trigger, the nr.e ligent navigation commences with instructions to the car park approach road, or to a particular approach road if there is a plurality. A second trigger generates instructions for which car park (when there is a plurality), a third trigger generates the route to a region of the allocated car park (within which there may be further triggers depending on complexity) and a fourth trigger generates allocation of parking oay. The triggers are usually spaced apart in time, but simultaneous triggering andi route assignments are also possible. The system also updates routing, and the Queue Ticket, depending on other monitored events and vehicles.

Figures 6 and 7 illustrate schematica ly two relatively simple examples of proximity and Queue Ticket triggering layouts.

Figure 6 illustrates a simple single park 60 that can be approached from two directions by approach roads 6' and 62. Queue Tickets are assigned when participating vehicles cross into the Parking Region marked by dotted boundary line 63. The Parking Region boundary may be determined by a variety of methods, including manually assigned, time to arrival, traffic management zoning or distance to arrival. For simplicity Figure 6 shows a symmetrical layout but in practice the Parking Region boundary may be asymmetric to cater for example for road layout, predominance of particular approach routes or approach route general speed. It may also be possible to modify the boundary position as a consequence of real time data as well as advance planning

Once a Queue Ticket is assigned, (box 51 in Figure 5) the location and movement of that vehicle will be considered as part of the optimal parking ordering of vehicles and the corresponding instructions generated. Some of the ordering processes associated with a Queue Ticket may provisionally take place or be updated before it is necessary to send individual row or bay instructions.

Figure 7 shows a slightly more comptex vehicle park layout, again having two approach roads, each of which lead to two car parks 71 and 72. Significant events and venues are likely have a multiplicity of car parks within the parking region and a range of approach roads, leading to an even larger multiplicity of Thresholds via which vehicles enter the Parking Region. Thresholds, approach roads, current estimated travel time and other information will all be used along with the Queue Ticket to determine the most appropr ate car park at the point of entering the Parking Region.

Once in the Parking Region the on-board navigation will update to the desired car park. As the vehicle then approaches the specific car park it has been assigned, the next trigger will update the Queue Ticket and start the Intelligent and Connected Parking process represented by box 53. It is possible for navigation and parking triggers to be at the same point in time or separated, depending on the layout and complexity.

At commencement of the intelligent parking operation each queue ticket is assigned a parking bay (or set of parking bays proximate one another) and the satellite navigation updates again to provide directions to this specific parking bay location.

The parking system continues to compare the current geolocation information of the participating vehicles, the planned perking structure and its relative performance against modelled predictions and consider re-ordering the queue tickets, for example if there has been diversion, Tail Back or if a more optimal parking pattern is identified. This check is illustrated by the loop 54 in Figure 5. At around this stage and when vehicle type is amongst the known data, this is also analysed and if it indicates a very long manoeuvre time (e.g., long vehicle) a better parking solution with reduced average parking times may be to reduce the number of vehicles in the set sent down a particular row at the same time.

As indicated in the bottom loop 55 in Figure 5, the application will sequentially check to see if a vehicle is parked. Once it has, the parking location will be recorded against the known parking bays and :he vehicle will be removed from the entire system for the completion of the Event Parking.

The application will now be idle for that vehicle. It may be reinitiated if there is guidance for departure.

Figure 8 schematically illustrates the in-vehicle messaging explaining where to park, and Figures 9A and 8B illustrate a display. Additional guidance such as instructing drivers to park facing forwards rather than by reversing into parking spaces is suggested as it is quicker with less manoeuvring. Other instructions, such as when bays are accessible from more than one side may also be given. Default ‘park anywhere near’ instructions may also be sent in some circumstances.

Some car parks may be operated such that entry is only allowed to participating vehicles. Others may operate on incentives such as much reduced parking charges for participating vehicles, and compliance, to attract a predominance of participating vehicles. The system may also have access to pre-booked parking information in order to assist with strategies and simulations.

It must generally be assumed that there will always be instances where for some reason a vehicle fails to pa*k as guided or there are non-participating vehicles that intervene. Strategies may be devised to alleviate this interference. One strategy, as already explained, is the allocation of a set of parking bays so that if a oay has been occupied by a non-participating vehicle and hence not registered as occupied in the system, there will be adjacent allocated bays that can be used. Drivers themselves in a participating car may also avai themselves of parking in a different location either as a ‘rogue’ event or due to unregistered bay occupation. When that happens the bay actually used will become registered. The system may also check for repeated avoidance of parking in allocated bays, which might come about due to bad or overlapping parking in adjacent bays, ether obstacles, or presence of a nonparticipating vehicle and after two or th'ee ‘avoidance’ events mark the bay as occupied status. Parking bays may also be provided with occupancy sensing that can be fed to the system. This is particularly likely with electric charging bays or other business models such as premium parking.

As the system monitors parking and bay choices it will be possible to detect when an event such as rogue parking happens for no good excuse, such event can also be fed into the parking fee system.

In operation and depending on the size and layout of a car park, there may come a level of occupancy when the system can no longer be accurate, and it is more efficient to switch to free parking or a recommendation to take the first available bay.

An important part of the database is a mapping of the carpark spatial layout so that where participating vehicles are at a relevant moment with regard to approach roads, any branching lanes, rows or parking bays are known by the system. This mapping may be recorded by any suitable programming or data means, for example by data representing a series of nodes and line segments which is briefly described below.

Figure 10 shows a schematic illustration of a node and line segment diagram representing stored carpark data. In data terms the nodes will be stored for example by an index reference numbered 1 , 2, 3, and so on, separated by sequentially indexed line segments A, B, C and so on.

At each node car parking spaces may be assigned or there may be sub paths also having nodes from which parking is assigned or which define further sub-paths.

Any number of bays may be assigned to a node depending on the layout, but each parking bay is assigned to only one node. Thus, in Figure 10 the second and third nodes (counting from the left) are each associated with a parking bay represented by rectangles 101. The fifth node has two associated bays.

It will be appreciated that a large park will tend have a more complex structure and may have several dividing pathways. Branching and sub-paths are shown in Figure 11 between line segments D and E and the sub-path 5BA itself also branches. Suitable indexing can be used to feedback sub-path data to its originating node. In Figure 12 the sub-paths from the fifth node are generally referenced 5A and 5B with nodes on that sub-path adding a numerical identifier (5A1 , 5A2 and 5B1 5B2 etc.) and the line segments having an alphabetical identifier (5AA, 5AB, and 5BA, 5BB etc.). Those nodes in turn are then associated with parking bays or may themselves branch further. Nodes on pathways may be placed at regular intervals and not all are ‘active’ in the sense of having parking or a pathway extending from them. They may be used for other purposes such as monitoring proximity of vehicles.

Other data will also be stored depending on the characteristics of the parking bays such as whether they are s ngle entry or double entry and the format in which parking is to happen, i.e., parallel or in line (as along the side of a path or road) or at an angle (usually perpendicular to the path or road). Figure 12 schematically illustrates angle and parallel parking and also single and double entry bays. ‘Single entry’ means where the parking can only happen from one end or side, usual y due to a barrier. ‘Double entry’ means the bay can be approached from both ends or sides although data then has to keep record of when an already parked vehicle blocks a common entry.

Figure 13 illustrates the node and segment diagram with reference to the car park layout shown schematically in Figures 1 to 4. In each of the top row D and bottom row A of Figure 13 there is a row of nodes each of which is associated with a single parking bay of an ‘angle’ tyoe. If these were single entry parking bays (the usual arrangement envisaged and expected from the node arrangement in the Figure) parking into those spaces would only be accessible, respectively, from rows D and A. However, if there were ro physical barrier, it would be possible to drive through (assuming no occupancy) and park in a bay that is respectively off Row C or B. If such an arrangement were required both the spaces would usually be indexed to the same node.

In rows C and D each node is shown indexed to two parking bays, one in each of the upper and lower blocks of spaces. Again, it is envisaged that these are single entry bays, but alternative double entry arrangements are also possible. Double entry or even end 1c end stacks of vehicles could be accommodated in other parking configurations. This may be useful for freight or lorry parks for example, where reversing is less desirable and sequencing of exit is controlled.

In the preferred embodiment of the invention the distribution of vehicles into rows has been made so that oveTaking in rows or waiting to park has predominantly been eliminated. This happens due to participating vehicles largely obeying instructions and thus disrupting the behaviour patterns that cause delay.

It may also be possible to use other formats of direction depending upon car park layout. In some instances, i: may be suffic ent merely to disrupt the parking behaviour patterns by distribution patterns that follow a random pattern or a random initial pattern with additiona criteria overlaid. With random patterns the majority of vehicles may still go to spaces separated ^_ 'rom other parking vehicles, although more interactions will occur. Such an arrangement might be used at times of lower throughput or when large parks are still relatively empty and driver actions to select alternative spaces can be tolerated. It may also be assisted by allocation of large sets of bays.

Minimising behavioural clustering and dither time each contribute to reducing the time to park. The spacing apart of parking cars also makes it easier and possibly safer for occupants of the cars to get out as there should not usually be manoeuvring going on immediately adjacent. The spacing apart may also contribute to less dither time used by drivers having to avoid pedestrians disembarking from adjacent vehicles.