Aircraft

Abstract

Whilst in-flight, the location of an aircraft 100 is controlled according to a flight path which is selected on the aircraft as follows. The aircraft’s flight plan (103,Fig.1), indicating a route (201

Description

Technical Field The present invention concerns aircraft control. More particularly, but not exclusively, this invention concerns a method for controlling an aircraft on the basis of a cost function. The invention also concerns an aircraft control system arranged to perform such a method. Background Aircraft are typically flown according to a flight plan. Such a flight plan defines a route through an airspace from an origin to a destination. The flight plan is generally decided in advance of the start of the flight and takes account of known air corridors and planned air traffic. Thus, an aircraft following an approved flight plan should, in the absence of any unexpected external factors, be able to safely follow the flight plan from the origin to the destination. However, external factors can arise that prevent the aircraft from safely following the flight plan. Weather conditions (for example, wind) can result in the aircraft being blown off course. Certain types of aircraft (for example, hang gliders) are not required to file flight plans and therefore cannot be anticipated during the flight planning process, meaning that they can present as obstacles to following the flight plan. In such situations, the pilot of an aircraft takes action to avoid any obstacles and return the aircraft to the flight plan in a safe manner. Aircraft (both piloted and autonomous) typically attempt to follow their flight plan as closely as practically possible, but may be driven by external factors (for example, wind or other aircraft) to divert from it. In particular, autonomous aircraft are typically programmed to follow their flight path unless it is absolutely necessary to divert from the flight plan in order to ensure the safety of the aircraft. In such cases, it is often necessary for a pilot to remotely take control of the autonomous aircraft in order to avoid any hazards. Aircraft often incorporate an Airborne Collision Avoidance System (AC AS). AC AS is a last-resort safety measure which operates independently of the aircraft’s normal navigation and flight control systems (including, for example, an autopilot system). In piloted aircraft, the system will typically provides warnings and recommendations to the pilot to assist them in avoiding mid-air collision. In some implementations, the system automatically intervenes, overriding ordinary flight control systems in order to avoid a mid-air collision. When doing so, AC AS gives no regard to the flight plan or to the aircraft’s mission; its only concern is avoiding a mid-air collision. As ACAS is intended as a safety-critical last-resort collision avoidance measure, it is configured to only take over in response to a significant risk of a mid-air collision. Thus, it would be inappropriate to rely solely on ACAS to avoid mid-air collisions, as it would result in the aircraft taking disproportionately large responses at a relatively late stage, even where the risk of mid-air collision may have been entirely avoided had the aircraft taken a much more minor action (for example, a minor heading adjustment) at an earlier point in time. Furthermore, as ACAS operates with no regard for the aircraft’s flight plan or mission, the responses it takes can divert the aircraft from the flight plan to such an extent that the flight plan must be abandoned. Additionally, although an approved flight plan may have been deconflicted with other flight plans, action by ACAS may cause the aircraft to diverge from its flight plan in sauch a way that it conflicts with the flight plan of a second aircraft, increasing the risk of collision with that second aircraft. This could also result in a cascade failure, whereby intervention by ACAS on a first aircraft causes a situation to arise which requires intervention by ACAS on one or more further aircraft, which may in turn result in intervention by ACAS on yet further aircraft. The present invention seeks to mitigate the above-mentioned problems. Alternatively or additionally, the present invention seeks to provide an improved method of controlling an aircraft. Summary The present invention provides, according to a first aspect, a method of controlling an aircraft. The method comprises, on the aircraft whilst in-flight: retrieving a flight plan for the aircraft, the flight plan indicating a route through an airspace from an origin to a destination; determining a current location of the aircraft; evaluating, for each of a plurality of possible flight paths from the determined current location, a cost function to calculate a flight path cost metric associated with the possible flight path, wherein the cost function is based on distance from the retrieved flight plan; selecting, using the determined flight path cost metrics, one of the plurality of possible flight paths; and controlling the location of the aircraft on the basis of the identified flight path. It will be appreciated by the skilled person that, whilst a flight plan is typically designed to account for the flight dynamics of an aircraft (including, for example, its maximum speed and turn radius), the ability of the aircraft to follow the flight plan can be affected by external factors. For example, weather conditions (including, for example, wind speeds) can, in the absence of action to compensate for those weather condition, result in an aircraft being unintentionally diverted away from its flight plan. It is therefore important that any autonomous (or semi-autonomous) aircraft control system is capable of adapting to such external factors. It will be appreciated by the skilled person that a “semi-autonomous” aircraft is one which operates by a combination of autonomous control and piloted control. Such aircraft may, for example, require a pilot to take control of the aircraft is certain circumstances. Alternatively, the aircraft may ordinarily be piloted remotely, but be capable of autonomous flight (for example, limited autonomous flight) in the event that a connection between the aircraft and the remote pilot is lost or degraded. The method of the present invention, by evaluating a cost function for each of a plurality of possible flight paths and, on the basis of the resulting cost metrics, selecting one of the plurality of possible flight paths, enables an aircraft to perform autonomous tactical guidance and control. In particular, the cost function operates to provide an indication of the suitability of a possible flight path. This enables the aircraft to quantify and compare possible flight paths to determine a preferred course of action. Thus, the method of the present invention provides improved techniques for autonomous aircraft control. It may be that the flight plan comprises a plurality of waypoints. In such cases, it may be that each of the plurality of waypoints corresponds to a respective location within the airspace. It may be that the flight plan defines a volume of the airspace through which the aircraft is planned to fly. The volume may be defined by the plurality of waypoints. For example, the plurality of waypoints may define a path through the airspace which, in combination with a defined allowable deviation distance of the aircraft from the path, defines the volume of the airspace. Alternatively, the flight plan may directly define a volume of the airspace. It may be that the flight plan is associated with timing data defining a planned progress of the aircraft along the flight plan towards the destination over time. Where the flight plan comprises a plurality of waypoints, it may be that each of the waypoints is associated with a time at which the aircraft is planned to arrive at the waypoint. It may be that each of the waypoints is associated with a specific time at which the aircraft is planned to arrive at the waypoint. Alternatively, each of the waypoints may be associated with a time window during which the aircraft is planned to arrive at the waypoint. It may be that the flight plan is stored in a memory located on the aircraft. In such cases, retrieving the flight plan may comprise retrieving the flight plan from the memory. Thus, it may be that the retrieving of the flight plan is performed without any communication with any external computing resource. Alternatively, it may be that the method comprises retrieving the flight plan from a remote computing resource (for example, via a data link). In such cases, it may be that the aircraft comprises a transceiver configured to communicate with the remote computing resource (for example, via wireless communication link). It may be that the plurality of possible flight paths are predetermined relative to a current location of the aircraft. It may be that the plurality of possible flight paths are determined on the basis of possible flight dynamics of the aircraft. It may be that the flight path cost metric quantifies a measure of risk associated with the respective flight path. It may therefore be that the cost metric comprises a dimensionless quantity (i.e. that the cost metric has no units) that acquires meaning only by virtue of its relative magnitude compared to one or more other cost metrics. It will be appreciated by the skilled person that, as the cost metric is used to select one of a plurality of possible flight paths, all of the possible flight paths having respective cost metrics, there is no need for the cost metric to indicate any absolute quantity in isolation for it to perform its function. Alternatively, it may be that that the risk is quantified by the cost metric as a monetary value (for example, as is the case when quantifying insurance risk). It will be appreciated by the skilled person that the cost function being based on distance from the retrieved flight plan does not preclude the cost function being additionally based on one or more further factors. Such further factors may, for example, include: a predicted distance of the aircraft from the retrieved flight plan, a predicted distance of the aircraft from the ground, a predicted distance of the aircraft from a further aircraft in the airspace, and a predicted progress of the aircraft towards the destination. The cost function may be based on any number of combination (for example, all) of these factors. It may be that the cost function comprises a cost field associated with distance from the retrieved flight plan. It may be that the cost function comprises a plurality of cost fields (for example, one associated with each of the factors listed above). By incorporating different cost fields into the cost function, it is possible to adapt the cost function to evaluate additional variables and information when determining cost metrics. Calculating a flight path cost metric for a possible flight path may comprise evaluating the cost function in respect of a plurality of locations along the possible flight path to determine a respective plurality of location cost metrics. It may be that the plurality of locations on the possible flight path comprise predicted future locations of the aircraft, in the event that it were to follow that possible flight path. Thus, calculating the flight path cost metric may comprise evaluating the cost function in respect of one or more locations in the airspace (for example, one or more predicted future locations of the aircraft). It may be that the calculating the flight path cost metric for a possible flight path comprises evaluating the cost function in respect of a plurality of locations along the possible flight path. It may be that the plurality of locations comprises at least three, at least five, at least ten, or at least fifteen locations. It may be that the plurality of locations for a given possible flight path are located at even distance intervals along the possible flight path. Alternatively, it may be that the plurality of locations may be located such that they are separated by smaller distance intervals at the beginning of the possible flight path (i.e. nearer to the current location of the aircraft) than at the end of the possible flight path. Where calculating a flight path cost metric comprises evaluating the cost function in respect of a plurality of locations to determine a respective plurality of location cost metrics, it may be that calculating the flight path cost metric comprises combining the plurality of location cost metrics to calculate the flight path cost metric. Combining the plurality of location cost metrics may comprise summing the location cost metrics. Alternatively, combining the plurality of location cost metrics may comprise calculating an average of the location cost metrics. It will be appreciated by the skilled person that the location cost metrics can also be combined to calculate a flight path cost metric in other ways. It may be that calculating the flight path cost metric for a possible flight path comprises integrating the cost function over the length of the possible flight path. Where the flight plan defines a volume of the airspace through which the aircraft is planned to fly, it may be that any location within that volume of airspace is considered to have a zero distance from the flight plan. Similarly, it may be that the distance to the flight plan comprises a distance to a nearest boundary of the volume of airspace. It will be appreciated by the skilled person that a location on the flight plan (i.e. one having zero distance from the flight plan) can be said to be in compliance with the flight plan. Likewise, a location off the flight plan (i.e. one having a non-zero distance from the flight plan) can be said to be not in compliance with the flight plan. The method may comprise selecting the possible flight path in the plurality having the lowest cost metric. It will be appreciated by the skilled person that the possible flight path having the lowest cost metric is the one assessed by the cost function as being the optimal one of the plurality of possible flight paths. Alternatively, it may be that the cost function operates as a filter to identify a predetermined number of the most suitable possible flight paths. For example, the selecting may comprise identifying flight paths in in the plurality having one of the ten lowest flight path cost metrics. Alternatively, it may be that the plurality of possible flight paths are filtered on the basis of their respective cost metrics to exclude those having a cost metric above a predetermined cost threshold. In either case, it may be that a single flight path is then selected on the basis of one or more other criteria. Controlling the location of the aircraft on the basis of the selected flight path may comprise controlling the aircraft to follow at least part of the selected flight plan. It will be appreciated by the skilled person that controlling the aircraft on the basis of the flight path does not require that the entire flight path be followed. It may be, for example, that only a portion of the flight path is followed before the aircraft selects a new flight path. This may happen, for example, due to the aircraft control system having subsequently repeated the steps of the method and selected an alternative flight plan. It will be appreciated that the data with which the cost function is evaluated may change over time. For example, it may be that the aircraft control system is only capable (for example, due to limited computational resources on the aircraft) of evaluating possible flight paths extending over a limited distance or period of time. The possible flight paths may therefore not extend to the destination. In such cases, movement of the aircraft will give a change in the portion of the airspace covered by the possible flight paths (i.e. such that the evaluation is subsequently based on different data) which may result in the selection of a different possible flight path. It may be that distance comprises a predicted distance of the aircraft from the retrieved flight plan (i.e. the cost function may be based on a predicted distance of the aircraft from the retrieved flight plan). Thus, the cost function may comprise a cost field associated with the predicted distance of the aircraft from the retrieved flight plan. The predicted distance of the aircraft from the retrieved flight plan may comprise a distance from the retrieved flight plan of a location in respect of which the cost function is being evaluated (for example, a predicted future location of the aircraft). It may be that the cost field associated with the predicted distance of the aircraft from the retrieved flight plan has a cost which increases with distance from the retrieved flight plan. It may be that the cost increases linearly as the distance increases. It may be that the cost increases exponentially as the distance increases. It may be that the cost increases as a step function. For example, it may be that the cost field attributes zero cost to locations in compliance with the flight plan and a non-zero cost to locations not in compliance with the flight plan. It may be that the cost field attributes a first cost to locations in compliance with the retrieved flight plan. It may be that the cost function attributes a second cost to locations not in compliance with the retrieved flight plan. In such cases, it may be that the first cost is a zero cost and the second cost is a non-zero cost. It may be that the second cost is the same for all locations not in compliance with the retrieved flight plan. It may be that the second cost is attributed only to locations having a distance from the retrieved flight plan of greater than zero and less than a predetermined distance threshold. It may be that the cost function attributes a third cost (for example, greater than the second cost) with locations having a distance from the retrieved flight plan of greater the predetermined distance threshold. Thus, the cost field may define a stepped function having multiple steps. It may be that the cost attributed by the cost field varies smoothly between the steps of the function (for example, such that there is a gradual transition between a first step and a second step as the distance from the flight plan increases). It may be that the flight plan defines one or more anticipated arrival times at one or more waypoints along the intended route. In such cases, it may be that the cost function is based on a predicted time of arrival of the aircraft at a point on the flight plan (for example, a waypoint defined by the flight plan). Thus, the cost function may comprise a cost field associated with predicted time of arrival of the aircraft at a point on the flight plan (for example, a waypoint defined by the flight plan). The cost function may be based on a difference between a predicted time of arrival at a point on the flight plan and an anticipated arrival time (for example, as defined by the flight plan) at that point. It may be that the point on the flight plan is not a waypoint defined in the flight plan. For example, it may be that an determining an anticipated arrival time comprises interpolating between the anticipated arrival times associated with two waypoints defined in the flight plan (for example, to determine an anticipated arrival time of an intermediate point). It may be that the cost function is based on a predicted risk of a collision between the aircraft and another object (for example, a static object). A static object may, for example, comprise the ground, a building, or other structure. It may be that the predicted risk of a collision when the aircraft is predicted to be in compliance with the flight plan is assumed to be zero. It may be that the predicted risk of a collision when the aircraft is predicted to be not in compliance with the flight plan is assumed to be non-zero. It will be appreciated by the skilled person that the predicted risk can be assumed to be zero when the aircraft is in compliance with the flight plan because avoidance of static objects is accounted for in the process of generating the flight plan. Thus, as long as the aircraft is in compliance with the flight plan, it can be assumed that there is a negligible risk of collision with static objects. It may be that the cost function is further based on a predicted distance of the aircraft from the ground. Thus, the cost function may comprise a cost field associated with the predicted distance of the aircraft from the ground. It will be appreciated by the skilled person that the “ground” in this context refers to the terrain beneath the airspace. Thus, the “ground” in this context may also refer to the surface of a body of water (for example, a lake or the sea). Basing the cost function on a predicted distance of the aircraft from the ground allows the aircraft to account for proximity to the ground (and the associated risks) when evaluating possible flight paths. It may be that the cost field associated with the predicted distance of the aircraft from the ground has a cost which increases with as the predicted distance decreases. It may be that locations above a certain distance from the ground (for example, associated with a predetermined minimum safe ground clearance) have a zero cost. It may be that the cost increases linearly as the distance decreases. It may be that the cost increases exponentially as the distance increases. It may be that the cost increases as a step function. It may be that the cost field attributes a first cost to locations at or above a predetermined ground distance threshold. It may be that the cost function attributes a second cost to locations below the predetermined ground distance threshold. In such cases, it may be that the first cost is a zero cost and the second cost is a non-zero cost. It may be that the second cost is the same for all below the predetermined ground distance threshold. Alternatively, the cost field may define a cost that escalates (for example, linearly or exponentially) with increasing proximity of the aircraft to the ground. The method may comprise obtaining topographical data for a region associated with (for example, beneath) the airspace. In such cases, the topographical data may be stored in a memory on the aircraft. Obtaining the topographical data may therefore comprise retrieving the data from the memory. Alternatively, it may be that the method comprises retrieving the topographical data from a remote computing resource (for example, via a data link). It may be that retrieving the topographical data comprises operating a sensor (for example, a radar or LIDAR sensor) to measure the terrain over which the aircraft is predicted to fly. Thus, obtaining the topographical data may comprise operating a sensor to measure the topographical data. The method may further comprise determining, based on the obtained topographical data and a predicted location of the aircraft, the predicted distance of the aircraft from the ground. Thus, it may be that the predicted distance of the aircraft from the ground is that at predicted location. It may be that the cost function is further based on a predicted distance of the aircraft from a further aircraft in the airspace. Thus, the cost function may comprise an obstacle distance cost field associated with the predicted distance of the aircraft from the further aircraft. The predicted distance of the aircraft from the further aircraft may comprise a distance from the further aircraft of a location in respect of which the cost function is being evaluated (for example, a predicted future location of the aircraft). It may be that the obstacle distance cost field associated with the predicted distance of the aircraft from the further aircraft has a cost which increases with proximity to the further aircraft. It may be that the cost increases linearly as the distance decreases. It may be that the cost increases exponentially as the distance decreases. It may be that the cost increases as a step function (for example, one having multiple steps as previously described in respect of other cost fields). It may be that the obstacle distance cost field attributes a first cost to locations in the airspace having a distance to the further aircraft greater than or equal to a predetermined obstacle distance threshold. It may be that the obstacle distance cost field attributes a second cost to locations in the airspace having a distance to the further aircraft less than the predetermined obstacle distance threshold. In such cases, it may be that the first cost is a zero cost and the second cost is a non-zero cost. It may be that the cost attributed by the obstacle distance cost field increases (for example, linearly or exponentially) with increasing proximity to the further aircraft. It may be that the cost function is further based on a risk associated with the aircraft crashing at a predicted future location of the aircraft. It may be that the method comprises retrieving (for example, from memory on the aircraft) data defining one or more relatively high risk regions of the airspace. It may be that the cost function is further based on a predicted distance of the aircraft from the one or more relatively high risk regions. Thus, the cost function may comprise a risk cost field associated with the predicted distance of the aircraft from the one or more relatively high risk regions. It may be that the risk cost field attributes increasing cost as proximity to the one or more relatively high risk regions increases. It may be that the method comprises retrieving (for example, from memory on the aircraft) data defining one or more relatively low risk regions of the airspace. The risk cost field may be further associated with the predicted distance of the aircraft from the one or more relatively low risk regions. It may be that the risk cost field attributes decreasing cost as the distance from the one or more relatively high risk regions increases. It may be that the one or more relatively high risk regions correspond to locations in which the likely consequences of the aircraft crashing would be more severe. It may be that the one or more relatively low risk regions correspond to locations in which the likely consequences of the aircraft crashing would be less severe. For example, it may be that developed areas are associated with relatively high risk regions, whilst regions located over the sea are associated with a relatively low risk. It may be that the aircraft comprises a sensor. The sensor may comprise one or more of a radar sensor, a LIDAR sensor, a visible light camera, and an infra-red camera. It will however be appreciated by the skilled person that other types of sensor may, alternatively or additionally, be used. The method may comprise operating the sensor to collect data characterising at least part of the airspace. In such cases, the collected data may indicate the presence of the further aircraft in the at least part of the airspace. Thus, the method may comprise detecting, using a sensor, a further aircraft in the airspace. The method may comprise predicting, on the basis of the collected data, a future location of the further aircraft. It may be that the method comprises determining a predicted distance of the aircraft from the predicted future location of the further aircraft. Predicting the future location of the further aircraft may comprise determining a current location of the further aircraft, determining a current direction and speed of travel of the further aircraft, and, on the basis of the determined current location and direction and speed of travel, predicting an expected flight path of the further aircraft. It may be that the collected data indicates the presence of a plurality of further aircraft in the at least part of the airspace. Thus, the method may comprise detecting, using a sensor, a plurality of further aircraft in the airspace. In such cases, it may be that the method comprises predicting, on the basis of the collected data, respective future locations of each the plurality of further aircraft. It may be that the method comprises determining a plurality of predicted distances of the aircraft from the respective predicted future locations of each of the of plurality of further aircraft. It may be that the cost function is based on the plurality of predicted distances. It may be that the cost function is based on a smallest of the plurality of predicted distances. It may be that the cost function is based on an average value of the plurality of predicted distances. It will be appreciated by the skilled person that these options are not mutually exclusive (i.e. the cost function may be based on both a smallest of the plurality of predicted distances and an average value of the plurality of predicted distances). Use of a cost function based on a predicted distance of the aircraft from a further aircraft in the airspace can allow the aircraft control system to account for the positions and flight paths of other aircraft (including, for example, aircraft not required to file flight plans in advance of their flights) when controlling the location of the aircraft. Thus, such a cost function can enable the aircraft to autonomously avoid other aircraft in the airspace. It may be that the cost function is further based on a predicted progress of the aircraft towards the destination. Thus, the cost function may comprise a cost field associated with the predicted progress of the aircraft towards the destination. It will be appreciated that the progress of the aircraft towards its destination relates to the proportion of the flight plan which the aircraft can be considered to have completed. Thus, it may be that the progress comprises a distance travelled by the aircraft along the flight plan. Where the aircraft is not in compliance with the flight plan (and thus, can be said to be “off the flight plan”), it may be that the progress comprises a distance along the flight plan of a nearest point on the flight plan to the aircraft. Alternatively, where the aircraft is not in compliance with the flight plan, it may be that the progress comprises a distance travelled by the aircraft along the flight plan up to its divergence from the flight plan. It may be that the progress comprises a distance (for example, a straight-line distance) of the aircraft from the destination. Use of a cost function based on a predicted progress of the aircraft towards the destination can provide an incentive for the system to progress along the flight plan. It will be appreciated by the skilled person that, where the cost function characterises the risk associated with locations within the airspace, the lowest risk course of action is for the aircraft to remain on the ground. Thus, use of a cost function based on a predicted progress of the aircraft towards the destination can help to prevent a situation arising in which the aircraft control excessively prioritises safety to the detriment of performing its intended purpose (for example, by simply refusing to take off). It will however be appreciated by the skilled person that such a cost field is not essential in all embodiments of the invention. For example, it may be that the cost function operates on the assumption that the aircraft must take off. Similarly, it may be that the cost function incorporates one or more other cost fields that operate to incentivise the completion of the aircraft’s mission. It may be that the cost function is further based on one of more flight dynamics of the aircraft. It may be that the cost function is further based on a manoeuvre cost of the possible flight path. Thus, the cost function may comprise a cost field associated with the manoeuvre cost of the possible flight path. It may be that the manoeuvre cost is associated with a severity of the manoeuvre defined by the possible flight path. Thus, the manoeuvre cost may be based on one or more of: a required bank angle of the aircraft, a turn radius of the possible flight path, a climb rate required by the possible flight path, a descent rate required by the possible flight path, and a flight speed required by the possible flight path. It may be that a flight path defining a more achievable or less severe manoeuvre is associated with a lower manoeuvre cost. It may be that at least one of the plurality of possible flight paths terminates at an end point not on the flight plan. It will be appreciated by the skilled person that the evaluation of possible flight paths is computationally intensive, and that the computational resources to perform the evaluation increases as the length of the flight path increases. As the computational resource available on an aircraft is limited, it may be that the maximum length of a possible flight path to be evaluated is restricted in order to ensure that the evaluation can be performed sufficiently quickly and / or repeated sufficiently frequently. Thus, the aircraft can be said to have a prediction horizon (i.e. a distance or time beyond which the aircraft does not predict its future behaviour). It will be appreciated by the skilled person that, where a possible flight path terminates at an end point not on the flight plan, there will typically be additional cost associated with the aircraft returning to the flight plan from that end point. Thus, in such cases, evaluating the cost function for the at least one of the plurality of possible flight paths may comprise estimating an additional cost associated with returning the aircraft to the flight path from the end point. It may be that estimating the additional cost comprises estimating the cost associated with a return path from the end point back to the flight plan. It may be that the return path is a straight line from the end point to a nearest point of the flight plan. Alternatively, the return path may be curved to account for achievable flight dynamics (for example, a turn radius) of the aircraft. It may be that the aircraft comprises an unmanned aircraft system. It may be that the aircraft comprises an unmanned aircraft system having a mass of greater than 250g (for example, greater than 500g, greater than 1kg, greater than 5kg, greater than 100kg, or greater than 1000kg). It may be that the aircraft comprises an unmanned aircraft system having a wingspan of at least 50cm (for example, at least 100cm, at least 500cm, at least Im, at least 3m, at least 10m, or at least 25m). A second aspect of the invention provides a computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out a method according to the first aspect. A third aspect of the invention provides an aircraft control system comprising a data retrieval module, a tactical guidance module, and an autopilot module. The data retrieval module is configured to retrieve a flight plan for the aircraft, the flight plan indicating a route through an airspace from an origin to a destination, and determine a current location of the aircraft. The tactical guidance module is configured to evaluate, for each of a plurality of possible flight paths from the determined current location, a cost function to calculate a flight path cost metric associated with the possible flight path, wherein the cost function is based on distance from the retrieved flight plan, and select, using the determined flight path cost metrics, one of the plurality of possible flight paths. The autopilot module is configured to control the location of the aircraft on the basis of the identified flight path. A fourth aspect of the invention provides an aircraft comprising the aircraft control system of the third aspect. It will of course be appreciated that features described in relation to one aspect of the present invention may be incorporated into other aspects of the present invention. For example, the method of the invention may incorporate any of the features described with reference to the apparatus of the invention and vice versa. Description of the Drawings Embodiments of the present invention will now be described by way of example only with reference to the accompanying schematic drawings of which: Figure 1 shows a schematic view of an aircraft according to a first embodiment of the invention; Figure 2 shows a schematic view illustrating the aircraft of Figure 1 following an example flight plan; Figure 3 shows a schematic view illustrating the aircraft of Figure 1 evaluating possible flight paths; Figure 4 shows a schematic view of an aircraft according to a third embodiment of the invention; and Figure 5 shows a flow chart illustrating the steps of a method according to a fifth embodiment of the invention. Detailed Description Figure 1 shows a schematic view of an aircraft 100 according to a first embodiment of the invention. In this example embodiment, the aircraft 100 comprises a quadcopter-type unmanned aircraft system. However, it will be appreciated that aircraft according to other embodiments of the invention may be manned and / or of other types (for example, a fixed-wing aircraft). The aircraft 100 comprises a data retrieval module 101. The data retrieval module 101 is configured to retrieve a flight plan 103 for the aircraft 100 and generate a flight plan data 105 indicative of the flight plan 103. In this example embodiment, the data retrieval module 101 is configured to retrieve the flight plan 103 from a memory 107. In alternative embodiments, it may be that the data retrieval module 101 is configured to retrieve the flight plan 103 from a remote computing resource (for example, a remote server). In such embodiments, it may be that the aircraft comprises a transceiver and / or antenna configured to provide a communication link between the aircraft and the remote computing resource. Figure 2 shows a schematic view of an example of the aircraft 100 following an example flight plan. Figure 2 is illustrated from a plan view perspective, such that the aircraft 100 is viewed from above. The flight plan 103 comprises a plurality of waypoints 201a, 201b, 201c, 20 Id in an airspace 203. Whilst the example flight plan illustrated in Figure 2 consists of four waypoints, it will be appreciated by the skilled person that a flight plan may consist of any number of waypoints. Each of the plurality of waypoints 201a, 201b, 201c, 20Id is associated with a respective location within the airspace. The plurality of waypoints 201a, 201b, 201c, 20 Id together define a flight path 205 through the airspace, along which the aircraft 100 is planned to fly. Thus, the flight plan 103 indicates a route through the airspace 203 from an origin to a destination. The flight plan 103 also defines an Operational Intent Volume (OIV) for the aircraft 100. The OIV (the bounds of which are illustrated by dashed lines 207a, 207b) indicates the portion of the airspace within which the aircraft 100 is permitted to fly whilst following the flight path 205. It will be appreciated by the skilled person that the bounds of the OIV may be defined based one of more of (for example, a combination of all of): a total system error (i.e. the precision with which the aircraft 100 can maintain a flight path in the presence of positional measurement error, path plan definition error and pilot / autopilot path following precision), a flight manoeuvre envelope (i.e. space allowed for tactical manoeuvring of the aircraft 100 in the performance of its mission), and contingency volume (i.e. the space required to complete contingency protocols - for example, based on the time required for ACAS to trigger). The aircraft 100 can be considered to be in compliance with the flight plan 103 when located within the bounds of the OIV. When the aircraft 100 is located outside the bounds of the OIV, it can be considered to be in non-compliance with the flight plan 103. It will be appreciated by the skilled person that the aircraft 100 will not in practice be capable of perfectly following the flight path 205 (due to, for example, errors in GNSS measurements and / or weather conditions), and that therefore allowance must be made for minor deviations from the flight path 205. In this example embodiment, the OIV is defined as an allowable distance by which the aircraft 100 can depart from the flight path 205. However, it will be appreciated by the skilled person that in other embodiments the flight plan 103 may directly define the bounds of the OIV. Whilst Figure 2 illustrates the flight plan 103 and the OIV in plan view, it will be appreciated by the skilled person that the OIV also defines allowable deviations in altitude from the flight path 103. Thus, the flight plan 103 defines a volume of the airspace through which the aircraft is planned to fly. Returning to Figure 1, the aircraft 100 further comprises a Global Navigation Satellite System (GNSS) receiver 109. The GNSS receiver 109 is configured to determine a current location of the aircraft 100 and to generate location data 111 indicative of the determined current location. It will be appreciated by the skilled person that, in other embodiments, the aircraft may further comprise one or more of an inertial measurement unit (IMU), a magnetometer, a barometer, and a pitot-static system. In such cases, the data output by the GNSS receiver 109 may be combined with the inertial measurement unit (IMU) data, magnetometer data, barometer data and / or the a pitot-static system data to allow a more precise determination of the position of the aircraft 100. The aircraft 100 further comprises a tactical guidance module 113. The tactical guidance module 113 is configured to receive the flight plan data 105 and the location data 111. The tactical guidance module 113 operates to determine a flight path for the aircraft as described below with reference to Figure 3. Figure 3 shows a schematic view illustrating the process by which the tactical guidance module 113 determines a flight path for the aircraft 100. The tactical guidance module 113 is configured to assess a plurality of possible flight paths 301a, 301b, 301c, 301 d, 301e, 301 f, 301g. In this example, the tactical guidance module 113 is configured to assess seven possible flight paths. However, it will be appreciated by the skilled person that the tactical guidance module 113 can be configured to assess a greater or lesser number of possible flight paths. Furthermore, whilst Figure 3 illustrates possible flight paths including motion in two dimensions only (i.e. in horizontal directions only), it will be appreciated by the skilled person that, in reality, the plurality possible flight paths will also include flight paths including vertical motion (for example, in combination with horizontal motion) and changes in aircraft velocity (for example, in combination with one or both of horizontal and vertical motion). In this example embodiment, the tactical guidance module 113 is preprogrammed to assess a predetermined plurality of possible flight paths from its current location. The trajectory of each of the plurality of possible flight paths is predetermined, such that a given possible flight path will only vary in respect of its start location (i.e. in respect of the location of the aircraft 100 at the time that possible flight path is assessed). That is to say that the route defined by that given possible flight path relative to its start location does not change depending on the location of the aircraft 100 or the time at which the possible flight path is assessed. The tactical guidance module 113 is configured to assess a possible flight path by evaluating a cost function in respect of the flight path. Thus, the tactical guidance module 113 is configured to evaluate, for each of a plurality of possible flight paths from the determined current location, a cost function to calculate a flight path cost metric associated with the possible flight path. Evaluating the cost function in respect of a possible flight path comprises evaluating the cost function in respect of a plurality of locations along that possible flight path to determine a respective plurality of location cost metrics. Thus, the tactical guidance module 113 is configured to determine, for each of the plurality of possible flight paths, a respective plurality of location cost metrics. In this example, embodiment, the tactical guidance module 113 is configured to determine 20 location cost metrics for each possible flight path. In this example embodiment, the plurality of locations are evenly spaced along the length of the possible flight path. However, it will be appreciated by the skilled person that the plurality of locations may comprise a greater or lesser number of locations and that those locations need not necessarily be evenly distributed along length of the possible flight path. For example, the plurality of locations may be spaced at smaller intervals at the beginning of the possible flight path (i.e. nearer to the current location of the aircraft 100) than at the end of the possible flight path. Evaluating the cost function in respect of a possible flight path further comprises combining the plurality of location cost metrics associated with that possible flight path to calculate the respective flight path cost metric. In this example embodiment, the location cost metrics are combined by summing the location cost metrics. The cost function includes a cost field (a “flight plan distance cost field”) which is based on a distance from the flight plan 103 of the location in respect of which the cost function is being evaluated. It will be appreciated by the skilled person that the locations in respect of which the cost function is evaluated are all possible future locations of the aircraft 100, in the event that it follows the associated possible flight path. Thus, the cost function can be said to be based on a predicted distance of the aircraft 100 from the retrieved flight plan 103. The flight plan distance cost field attributes a first cost to locations in compliance with the retrieved flight plan 103. The flight plan distance cost field attributes a second cost to locations not in compliance with the retrieved flight plan 103. In this example embodiment, the first cost is a zero cost and the second cost is a non-zero cost. The second cost is the same for all locations not in compliance with the retrieved flight plan 103. Thus, the flight plan distance cost field attributes zero cost to locations within the OIV and a non-zero cost to locations outside of the OIV. It will be appreciated by the skilled person that other embodiments of the invention may apply different implementations of the flight plan distance cost field. For example, in such other embodiments, it may be that the cost associated with locations outside of the OIV by the flight plan distance cost field varies linearly according to the distance of the location from the flight plan 103 (for example, such that the cost associated with a location outside the OIV increases as the distance from the OIV increases). In this example embodiment, the cost function also includes a cost field (a “mission progress cost field”) which is based on a predicted progress of the aircraft 100 towards the destination. The mission progress cost field attributes to a location a cost which depends on the distance of the location from the destination, with locations further from the destination having a higher associated cost than those closer to the destination. Thus, the mission progress cost field operates to incentivise the aircraft to travel towards the destination. However, in alternative embodiments this functionality may be include in the flight plan distance cost field. For example, it may be that the plurality of waypoints include timing data (for example, each waypoint may be associated with a respective time period during which the aircraft 100 is expected to arrive at the waypoint) which serve to adapt the volume of space defined by the flight plan 103 so as to incentivise progress of the aircraft 100 towards the destination. Because of the limited computing resource available on the aircraft 100, the length of each of the possible flight paths is constrained to a maximum length that is significantly less than that of the flight plan 103. As a result, each of the plurality of possible flight paths will, in most cases (for example, when not close to the destination), terminate at an end point that is not at the destination. Furthermore, it will generally be the case that a number of the possible flight paths terminate at an end point that is not on the flight plan 103. The cost function therefore also includes a cost field (a “return to flight plan conformance cost field”) which comprises an estimate of an additional cost associated with returning the aircraft 100 to the flight path 103 from the end point. Where a possible flight path terminates at a location on the flight plan 103, the return to flight plan conformance cost field attributes zero additional cost. Where a possible flight path terminates at a location that is not on the flight plan 103, the return to flight plan conformance cost field attributes an additional cost. The additional cost is determined on the basis of a straight-line distance from the end point to a nearest point on the flight path 103, with greater distances being associated with higher additional cost. Thus, the tactical guidance module 113 is configured to determine, for each of the plurality of possible flight paths, a flight path cost metric that indicates the appropriateness (according to the priorities defined by the cost function) of that possible flight. The tactical guidance module 113 is further configured to select, on the basis of the determined flight path cost metrics, one of the plurality of possible flight paths. In this example embodiment, the tactical guidance module 113 is configured to select the possible flight path in the plurality which has the lowest cost metric. The tactical guidance module 113 outputs guidance data 115 defining the selected flight path for use in controlling the location of the aircraft 100. The aircraft 100 further comprises an autopilot module 117. The autopilot module 117 is configured to control the location of the aircraft 100 (for example, by actuating one or more control surfaces on the aircraft 100) on the basis of the identified flight path. Thus, the aircraft 100 operates to autonomously assess a number of possible flight paths, select one of the possible flight paths, and control the aircraft 100 to follow the selected flight path. The aircraft 100 is configured to periodically repeat the steps of assessing a plurality of possible flight paths, selecting one of the plurality of possible flight paths, and controlling the aircraft 100 to follow the selected flight path over the course of the aircraft’s mission. In this example embodiment, the aircraft 100 is configured to repeat these steps once every second. However, it will be appreciated by the skilled person that the steps may alternatively be repeated at other frequencies. Because the aircraft 100 is continually evaluating and selecting a flight path, it is likely that the aircraft will in operation follow only a portion of a flight path before selecting a new flight path. It will be appreciated by the skilled person that, as the aircraft moves through the airspace, the locations in respect of which the cost function is evaluated will change, and therefore which of the possible flight paths is associated with the lowest flight path cost metric will also change. According to a second embodiment of the invention, there is provided an aircraft as described in respect of the first embodiment but for the following features. The data retrieval module 101 is further configured to obtain topographical data for a region associated with the airspace. The topographical data defines the shape and height of the terrain beneath the airspace. The tactical guidance module 113 is as described in respect of the first embodiment, but is further configured to receive the topographical data for use in evaluating the cost function. The cost function further includes a cost field (a “ground distance cost field”) which is based on the predicted distance of the aircraft from the ground. The tactical guidance module 113 is configured to determine, for each of the plurality of locations, a distance of the location from the ground. The distance is determined on the basis of the obtained topographical data. As previously discussed, as the plurality of locations are possible future locations of the aircraft 100, the tactical guidance module 113 can be said configured to determine a predicted distance of the aircraft 100 from the ground. Thus, the cost function can be said to be based on the predicted distance of the aircraft 100 from the ground. The ground distance cost field attributes a cost to a location on the basis of a distance of the location from the ground. In this case, the ground distance cost field attributes a first cost to locations having a distance to the ground at or above a predetermined ground distance threshold (for example, determined by safe operating limits). The ground distance cost field attributes a second cost to locations having a distance to the ground below the predetermined ground distance threshold. In this example embodiment, the first cost is a zero cost and the second cost is a non-zero cost. The second cost is the same for all locations having a distance to the ground below the predetermined ground distance threshold. Thus, the ground distance cost field attributes zero cost to locations at or above the predetermined ground distance threshold and a non-zero cost to locations having a distance to the ground below the predetermined ground distance threshold. It will be appreciated by the skilled person that other embodiments of the invention may apply different implementations of the ground distance cost field. For example, in such other embodiments, it may be that the cost attributed by the ground distance cost field to locations having a distance to the ground below the predetermined ground distance threshold cost field varies linearly according to the distance of the location from the ground (for example, such that the cost associated with a location increases as the distance from the ground decreases). Figure 4 shows an aircraft 400 according to a third embodiment of the invention. The aircraft 400 is the same as the aircraft 100 of the second embodiment apart from the following features. The aircraft 400 further comprises a sensor 419. The sensor 419 is configured to collect data characterising at least part of the airspace. In particular, the sensor 419 is configured to collect data indicating the presence of a further aircraft in the at least part of the airspace. Thus, the sensor 419 is configured to detect a further aircraft in the airspace. The sensor 419 is configured to generate sensor data 421 indicative of the detected further aircraft. The tactical guidance module 413 is as described in respect of the first embodiment, but is further configured to receive the sensor data 421 for use in evaluating the cost function. The cost function further includes a cost field (a “obstacle distance cost field”) which is based on a predicted distance of the aircraft 400 from the one or more further aircraft. As previously discussed, as the plurality of locations are possible future locations of the aircraft 400, the tactical guidance module 413 can be said configured to determine a predicted distance of the aircraft 400 from the further aircraft. Thus, the cost function can be said to be based on the predicted distance of the aircraft 400 from the further aircraft. It will be appreciated by the skilled person that, as the plurality of locations relate to possible locations of the aircraft 400 at future points in time, it is necessary to predict the locations of the further aircraft at that same point in time. Thus, the tactical guidance module 413 is configured to predict, on the basis of the collected data, a future location of the further aircraft. Predicting the future location comprises determining a current location, direction of travel, and speed of the further aircraft and, on the basis of the determined current location, and direction of travel, and speed, predicting an expected flight path of the further aircraft. It is then possible to predict a future location of the further aircraft at the time at which the aircraft 400 is expected to be at the location in respect of which the cost function is being evaluated. The tactical guidance module 413 can then determine the distance between the location in respect of which the cost function is being evaluated and the predicted location of the further aircraft to determine the predicted distance. According to a fourth embodiment of the invention, there is provided an aircraft as described in respect of the first embodiment but for the following features. The trajectory of the aircraft is defined as a curve through three-dimensional space indexed by time; mathematically this is a function u:[0,T]—> R3. The over-riding objective of the tactical guidance module is to minimise the following equation: fT (p[u] = q I ^(uCO.tjdt + c202(u(T),T) + c303(u(O), 0) Jq The first term of the equation represents the path costs (i.e. the cost integral for the possible flight path). The second term of the equation represents additional cost associated with returning to the flight path from the end point. This second term may therefore be zero in some cases (i.e. where the possible flight path terminates on the flight plan). The third term of the equation represents the initial cost. The path costs are calculated using the following equations: 1 , V        -^(OI-MjCOX / Mt)-G;(t)|+Mz(t)\\ Pi\t) — 1 + ~I errI------------==------------I — err ------------==------------I I 2\ \ 7¾ J \ 7¾ J J fT ^MISSION M = C I VP1(O2 + PzCO2 + Jo where: erf denotes the error function; ut is the Ith component of the current position of the aircraft; Gt is the ith component of ideal position of the craft according to the original plan; Oi is the Ith component of the total system error; and Mt is the greatest lower bound on the deviation of the aircraft from G in the Ith component. The additional cost associated with returning to the flight path from the end point is calculated using the following equation: E[C] = J E[Rc]dt Thus, the additional cost is determined by integrating the cost over a straight-line path back to the flight plan with respect to the time taken to return. Figure 5 shows a flow chart illustrating the steps of a method 500 according to a fifth embodiment of the invention. It may be that the aircraft comprises an unmanned aircraft system. It may be that the aircraft comprises a sensor (for example, for detecting obstacles). It may be that the method 500 is performed on the aircraft (for example, whilst the aircraft is in-flight). A first step, illustrated by item 501, of the method 500 comprises retrieving a flight plan for the aircraft. It may be that the flight plan indicates a route through an airspace from an origin to a destination; The flight plan may comprise a plurality of waypoints. In such cases, it may be that each of the plurality of waypoints corresponds to a respective location within the airspace. It may be that each of the waypoints is associated with a time period within which the aircraft is planned to arrive at the waypoint. It may be that the retrieved flight plan defines a volume of the airspace through which the aircraft is to fly. A second step, illustrated by item 503, of the method 500 comprises determining a current location of the aircraft. An optional third step, illustrated by item 505, of the method 500 comprises obtaining topographical data for a region associated with the airspace. An optional fourth step, illustrated by item 507, of the method 500 comprises determining, based on the obtained topographical data and a predicted location of the aircraft, a predicted distance of the aircraft from the ground. An optional fifth step, illustrated by item 509, of the method 500 comprises operating the sensor to collect data characterising at least part of the airspace, wherein the data indicates the presence of a further aircraft in at least part of the airspace. An optional sixth step, illustrated by item 511, of the method 500 comprises predicting, on the basis of the collected data, a future location of the further aircraft. It may be that predicting the future location of the further aircraft comprises determining a current location of the further aircraft. It may be that predicting the future location of the further aircraft comprises determining a current direction and speed of travel of the further aircraft. It may be that predicting the future location of the further aircraft comprises, on the basis of the determined current location, and direction and speed of travel, predicting an expected flight path of the further aircraft. An optional seventh step, illustrated by item 513, of the method 500 comprises determining a predicted distance of the aircraft from the predicted future location of the further aircraft. An eighth step, illustrated by item 515, of the method 500 comprises evaluating, for each of a plurality of possible flight paths from the determined current location, a cost function to calculate a flight path cost metric associated with the possible flight path. It may be that the cost function is based on distance from the retrieved flight plan. It may be that the cost function is based on a predicted distance of the aircraft from the retrieved flight plan. It may be that the cost function is based on a predicted distance of the aircraft from the ground. It may be that the cost function is based on a predicted distance of the aircraft from a further aircraft in the airspace. It may be that the cost function is based on a predicted progress of the aircraft towards the destination. It may be that the cost function is based on a progress along the flight plan of a nearest point on the flight plan to the aircraft. It may be that the cost function is based on a predicted risk of a collision between the aircraft and another object. In such cases, it may be that the predicted risk of a collision when the aircraft is predicted to be in compliance with the flight plan is assumed to be zero. It may be that the predicted risk of a collision when the aircraft is predicted to be not in compliance with the flight plan is assumed to be non-zero. Calculating a flight path cost metric for a possible flight path may comprise evaluating the cost function in respect of a plurality of locations along the possible flight path to determine a respective plurality of location cost metrics. In such cases, it may be that calculating the flight path cost metric comprises combining the plurality of location cost metrics to calculate the flight path cost metric. It may be that the cost function associates a first cost with locations in compliance with the retrieved flight plan. It may be that the cost function associates a second cost with locations not in compliance with the retrieved flight plan. The first cost may be a zero cost. The second cost may be a non-zero cost. The second cost may be the same for all locations not in compliance with the retrieved flight plan. It may be that at least one of the plurality of possible flight paths terminates at an end point not on the flight plan. In such cases, evaluating the cost function for the at least one of the plurality of possible flight paths may comprise estimating an additional cost associated with returning to the flight path from the end point. A ninth step, illustrated by item 517, of the method 500 comprises selecting, using the determined flight path cost metrics, one of the plurality of possible flight paths. It may be that the method comprises selecting the flight path in the plurality having the lowest cost metric. A tenth step, illustrated by item 519, of the method 500 comprises controlling the location of the aircraft on the basis of the identified flight path. Whilst the present invention has been described and illustrated with reference to particular embodiments, it will be appreciated by those of ordinary skill in the art that the invention lends itself to many different variations not specifically illustrated herein. By way of example only, certain possible variations will now be described. In embodiments described above, the flight plan is retrieved from a memory on the aircraft. However, in alternative embodiments, the flight plan is retrieved from a remote computing resource (for example, a remote server). In such embodiments, the aircraft may comprise a transceiver configured to facilitate communication between the data retrieval module and the remote computing resource. It may be that the aircraft further comprises an antenna (for example, for use in communicating with the remote computing resource). It may be that the transceiver is configured to communicate, by use of the antenna, over a wireless communication link. In embodiments described above the cost function is based on a predicted progress of the aircraft towards the destination. In those embodiments the predicted progress is based on a distance (for example, a straight-line distance) of the aircraft from the destination. However, in alternative embodiments, the predicted progress is determined based on the proportion of the flight plan that the aircraft has completed. In such cases, where the location in respect of which the cost function is being evaluated is not on the flight plan, the proportion may be determined on the basis of a nearest point on the flight plan to the aircraft. In general, whilst the embodiments above illustrate a number of different implementations of the cost function, it will be appreciated by the skilled person that other implementations are also possible. The cost function may, for example, incorporate any number and / or combination of the cost fields described above. It will also be appreciated by the skilled person that the magnitude of the costs associated with different cost fields can be varied in order to influence the costs attributed to different possible flight paths (and thereby also the behaviour of the aircraft 100). For example, increasing the costs associated with deviation from the flight plan 103 and decreasing the costs associated with progress towards the destination can result in the prioritisation of compliance with the flight plan 103 over progress towards the destination. It will be appreciated that aircraft 100, 400 may comprise one or more processors and / or memory (which may or may not be the same as memory 107). The processor and associated memory may be configured to perform one or more of the above-described functions of the aircraft 100, 400. Each device, module, component, machine, or function as described in relation to any of the examples described herein (for example, data retrieval module 101, GNSS receiver 109, sensor 419, tactical guidance module 113, 413, and autopilot module 117) may similarly comprise a processor or may be comprised in apparatus comprising a processor. One or more aspects of the embodiments described herein comprise processes performed by apparatus. In some examples, the apparatus comprises one or more processors configured to carry out these processes. In this regard, embodiments may be implemented at least in part by computer software stored in (non-transitory) memory and executable by the processor, or by hardware, or by a combination of tangibly stored software and hardware (and tangibly stored firmware). Embodiments also include computer programs, particularly computer programs on or in a carrier, adapted for putting the above-described embodiments into practice. The program may be in the form of non-transitory source code, object code, or in any other non-transitory form suitable for use in the implementation of processes according to embodiments. The carrier may be any entity or device capable of carrying the program, such as a RAM, a ROM, or an optical memory device, etc. The one or more processors of the aircraft 100, 400 may comprise a central processing unit (CPU). The one or more processors may comprise a graphics processing unit (GPU). The one or more processors may comprise one or more of a field programmable gate array (FPGA), a programmable logic device (PLD), or a complex programmable logic device (CPLD). The one or more processors may comprise an application specific integrated circuit (ASIC). It will be appreciated by the skilled person that many other types of device, in addition to the examples provided, may be used to provide the one or more processors. The one or more processors may comprise multiple co-located processors or multiple disparately located processors. Operations performed by the one or more processors may be carried out by one or more of hardware, firmware, and software. The one or more processors may comprise data storage. The data storage may comprise one or both of volatile and non-volatile memory. The data storage may comprise one or more of random access memory (RAM), read-only memory (ROM), a magnetic or optical disk and disk drive, or a solid-state drive (SSD). It will be appreciated by the skilled person that many other types of memory, in addition to the examples provided, may also be used. It will be appreciated by a person skilled in the art that the one or more processors may each comprise more, fewer and / or different components from those described. The techniques described herein may be implemented in software or hardware, or may be implemented using a combination of software and hardware. They may include configuring an apparatus to carry out and / or support any or all of techniques described herein. Although at least some aspects of the examples described herein with reference to the drawings comprise computer processes performed in processing systems or processors, examples described herein also extend to computer programs, for example computer programs on or in a carrier, adapted for putting the examples into practice. The carrier may be any entity or device capable of carrying the program. The carrier may comprise a computer readable storage media. Examples of tangible computer-readable storage media include, but are not limited to, an optical medium (e g., CD-ROM, DVD-ROM, or Blu-ray), flash memory card, floppy or hard disk or any other medium capable of storing computer-readable instructions such as firmware or microcode in at least one ROM or RAM or Programmable ROM (PROM) chips. Where in the foregoing description, integers or elements are mentioned which have known, obvious or foreseeable equivalents, then such equivalents are herein incorporated as if individually set forth. Reference should be made to the claims for determining the true scope of the present invention, which should be construed so as to encompass any such equivalents. It will also be appreciated by the reader that integers or features of the invention that are described as preferable, advantageous, convenient or the like are optional and do not limit the scope of the independent claims. Moreover, it is to be understood that such optional integers or features, whilst of possible benefit in some embodiments of the invention, may not be desirable, and may therefore be absent, in other embodiments.

Claims

1. A method of controlling an aircraft, the method comprising, on the aircraft whilst in-flight:retrieving a flight plan for the aircraft, the flight plan indicating a route through an airspace from an origin to a destination;determining a current location of the aircraft;evaluating, for each of a plurality of possible flight paths from the determined current location, a cost function to calculate a flight path cost metric associated with the possible flight path, wherein the cost function is based on distance from the retrieved flight plan;selecting, using the determined flight path cost metrics, one of the plurality of possible flight paths; andcontrolling the location of the aircraft on the basis of the selected flight path.

2. A method according to any preceding claim, wherein the distance is a predicted distance of the aircraft from the retrieved flight plan.

3. A method according to any preceding claim, wherein the cost function is further based on a predicted distance of the aircraft from the ground.

4. A method according to claim 3, wherein the method comprises:obtaining topographical data for a region associated with the airspace; and determining, based on the obtained topographical data and a predicted location of the aircraft, the predicted distance of the aircraft from the ground.

5. A method according to any preceding claim, wherein the cost function is further based on a predicted distance of the aircraft from a further aircraft in the airspace.

6. A method according to claim 5, wherein:the aircraft comprises a sensor; andthe method comprises operating the sensor to collect data characterising at least part of the airspace, wherein the data indicates the presence of the further aircraft in the at least part of the airspace.

7. A method according to claim 6, wherein the method comprises:predicting, on the basis of the collected data, a future location of the further aircraft; anddetermining a predicted distance of the aircraft from the predicted future location of the further aircraft.

8. A method according to claim 7, wherein predicting the future location of the further aircraft comprises:determining a current location of the further aircraft;determining a current direction and speed of travel of the further aircraft; and on the basis of the determined current location, and direction and speed of travel, predicting an expected flight path of the further aircraft.

9. A method according to any preceding claim, wherein the cost function is further based on a predicted progress of the aircraft towards the destination.

10. A method according to any preceding claim, wherein calculating a flight path cost metric for a possible flight path comprises:evaluating the cost function in respect of a plurality of locations along the possible flight path to determine a respective plurality of location cost metrics; andcombining the plurality of location cost metrics to calculate the flight path cost metric.

11. A method according to any preceding claim, wherein the method comprises selecting the possible flight path in the plurality having the lowest cost metric.

12. A method according to any preceding claim, wherein the flight plan comprises a plurality of waypoints, each of the plurality of waypoints corresponding to a respective location within the airspace.

13. A method according to claim 12, wherein each of the waypoints is associated with a time period within which the aircraft is planned to arrive at the waypoint.

14. A method according to any preceding claim, wherein the cost function is further based on a progress along the flight plan of a nearest point on the flight plan to the aircraft.

15. A method according to any preceding claim, wherein the cost function is further based on a predicted risk of a collision between the aircraft and another object.

16. A method according to claim 15, wherein:the predicted risk of a collision when the aircraft is predicted to be in compliance with the flight plan is assumed to be zero; andthe predicted risk of a collision when the aircraft is predicted to be not in compliance with the flight plan is assumed to be non-zero.

17. A method according to any preceding claim, wherein the retrieved flight plan defines a volume of the airspace through which the aircraft is to fly.

18. A method according to any preceding claim, wherein:the cost function associates a first cost with locations in compliance with the retrieved flight plan; andthe cost function associates a second cost with locations not in compliance with the retrieved flight plan.

19. A method according to claim 18, wherein the first cost is a zero cost and the second cost is a non-zero cost.

20. A method according to claim 18 or 19, wherein the second cost is the same for all locations not in compliance with the retrieved flight plan.

21. A method according to any preceding claim, wherein:at least one of the plurality of possible flight paths terminates at an end point not on the flight plan; andevaluating the cost function for the at least one of the plurality of possible flight paths comprises estimating an additional cost associated with returning to the flight path from the end point.

22. A method according to any preceding claim, wherein the aircraft comprises an unmanned aircraft system.

23. A computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out a method according to any of claims 1 to 22.

24. An aircraft control system comprising:a data retrieval module configured to retrieve a flight plan for the aircraft, the flight plan indicating a route through an airspace from an origin to a destination, and determine a current location of the aircraft;a tactical guidance module configured to evaluate, for each of a plurality of possible flight paths from the determined current location, a cost function to calculate a flight path cost metric associated with the possible flight path, wherein the cost function is based on distance from the retrieved flight plan, and select, using the determined flight path cost metrics, one of the plurality of possible flight paths; andan autopilot module configured to control the location of the aircraft on the basis of the selected flight path.

25. An aircraft comprising an aircraft control system according to claim 24.

Images