Actuator assembly

Abstract

An actuator assembly comprises first and second parts relatively moveable in a plane perpendicular to a primary axis, and a pari of actuating units each connected between the first and second parts to

Description

Field The present application relates to an actuator assembly with one or more actuating units, each of which includes a shape memory alloy (SMA) element. Background SMA actuator assemblies may be used in a variety of applications for moving a movable part relative to a support structure. For example, WO 2013 / 175197 Al describes a camera in which four SMA wires are arranged to move a lens element relative to an image sensor in a plane that is perpendicular to the optical axis of the lens element, thereby effecting optical image stabilization (OIS). WO 2010 / 029316 Al describes SMA actuator wires used to provide OIS in a camera by tilting a camera module. WO 2011 / 104518 Al describes an actuator assembly having eight SMA wires capable of effecting positional control of a movable element with multiple degrees of freedom. Typically, the range of movement (also known as "stroke") of such SMA actuator assemblies is limited by the maximum contraction of the SMA wires, and the actuating force is limited by the maximum force that can generated by the SMA wires. To increase the movement range or the actuating force, longer or thicker SMA actuator wires can be used, but this may be at the expense of increased cost, size and / or power, which may not be practical in miniature applications. WO 2022 / 084699 Al discloses an actuator assembly comprising at least one actuating unit (incorporating an SMA wire) that, on actuation, moves a movable part relative to the support structure. The actuating unit may be configured to amplify the movement range of the movable part, to amplify the actuating force acting on the movable part, or to re-direct the force applied by the SMA wire. Summary According to an aspect of the present invention, there is provided an actuator assembly comprising: first and second parts that are movable relative to each other in a movement plane that is perpendicular to a primary axis; and at least one pair of actuating units, each actuating unit connected between the first and second parts and arranged to apply a respective actuating force between the first and second parts, wherein one of each pair of actuating units comprises: a body portion; an SMA element connected between the body portion and the first part, and arranged, on actuation, to apply an input force to the body portion; a force-modifying element connected between the body portion and the first part and configured to modify the input force so as to give rise to the actuating force; and a coupling link connected between the body portion and the second part, wherein the coupling link is configured to transmit the actuating force from the body portion to the second part, and wherein the coupling link is compliant in a direction perpendicular to the direction of the actuating force; and wherein the other of each pair of actuating units comprises: a body portion; an SMA element connected between the body portion and the second part, and arranged, on actuation, to apply an input force to the body portion; a force-modifying element connected between the body portion and the second part and configured to modify the input force so as to give rise to the actuating force; and a coupling link connected between the body portion and the first part, wherein the coupling link is configured to transmit the actuating force from the body portion to the first part, and wherein the coupling link is compliant in a direction perpendicular to the direction of the actuating force. Further features are set out in the dependent claims. Brief Description of the Drawings Certain examples will now be described, byway of example only, with reference to the accompanying drawings, in which: Figures 1A-E are schematic cross-sectional views of different variations of a camera module incorporating an actuator assembly; Figure 2 is a schematic perspective view of the actuator assembly; Figures 3A and 3B are perspective and plan views of an actuating unit forming part of the actuator assembly; Figure 4 is a schematic plan view of an arrangement of four actuating units; Figures 5 is a schematic perspective view of an arrangement of eight actuating units; Figure 6A and B schematically depicts an embodiment of an actuator assembly including actuating units; Figure 7 schematically shows the height of the actuator assembly of Figure 6; Figures 8A-8B schematically depict an actuator assembly including actuating units; Figure 9 schematically depicts an actuating unit for use in the actuator assembly; Figure 10 schematically depicts an embodiment of an actuator assembly including actuating units; Figures 11A and 11B schematically depict an embodiment of an actuator assembly including actuating units; Figures 12A and 12B schematically depict another actuator assembly according to the present invention; and Figures 13A-13C schematically depict structural details of the actuator assembly. Figure 14 is a schematic plan view of an SMA actuator assembly, including four actuating units in two planes; Figure 15 is a schematic plan view of an SMA actuator assembly, including four actuating units in one plane; Figure 16 is a schematic perspective view of the SMA actuator assembly illustrated in Figure 4; Figure 17 is a schematic plan view of an SMA actuator assembly, including four actuating units; Figure 18 is a schematic diagram of a flexure and wire biasing the actuating unit against a bearing; Figure 19 is a schematic diagram of an actuating unit with an integrated crimp; Figures 20-25 are schematic plan views of alternative SMA actuator assemblies, each including four actuating units; and Figure 26 is a schematic diagram illustrating a wire on a pulley. Detailed description Camera module Figures 1A-E schematically shows different variations of an apparatus 1 incorporating an actuator assembly 2. The apparatus 1 is, for example, a camera module 1. Generally, the apparatus 1 is to be incorporated in a portable electronic device such as a smartphone. Thus, miniaturisation can be an important design criterion. Figure 2 schematically shows the actuator assembly 2. The actuator assembly 2 includes a support structure 10 and a movable part 20. The movable part 20 is movable relative to the support structure 10. When the actuator assembly 2 is included e.g. in the apparatus 1, the support structure 10 may be fixed relative to the main body of the apparatus 1. However, in general, the support structure 10 need not be stationary and may be movable relative to or within the apparatus 1. The actuator assembly 2 includes one or more actuating units 30. Each actuating unit 30 is configured to apply an actuating force to the movable part 20 capable of moving the movable part 20 relative to the support structure 10. The movable part 20 may be supported (i.e. suspended) on the support structure 10 exclusively by the actuating units 30. Alternatively, the actuator assembly 2 may include a bearing arrangement 40 that supports the movable part 20 on the support structure 10. The actuating units 30 and the bearing arrangement 40 may together support the movable part 20 on the support structure 10. The bearing arrangement 40 may have any suitable form for allowing movement of the movable part 20 with respect to the support structure 10 with one or more degrees of freedom (DOFs). The actuating units 30 and / or the bearing arrangement 40 may constrain, i.e. reduce or prevent, other DOFs of movement of the movable part 20 relative to the support structure 10. For this purpose, the bearing arrangement 40 may, for example, include one or more of the following bearings: a rolling bearing (such as a ball bearing), a flexure bearing (i.e. an arrangement of flexures or other resilient elements that guide movement), or a plain (i.e. sliding contact) bearing. A primary axis P can be defined with reference to the actuator assembly 2 and / or the support structure 10. The primary axis P may extend through the actuator assembly 2, e.g. through the centre of the actuator assembly 2. In some examples, the actuator assembly 2, the support structure 10 and / or the movable part 20 extends predominantly in a direction perpendicular to the primary axis P. In other words, the extent of the actuator assembly 2, the support structure 10 and / or the movable part 20 along the primary axis P is less than the extent thereof along any direction perpendicular to the primary axis P. Alternatively or additionally, the support structure 10 and / or movable part 20 may include a planar component that extends perpendicularly to the primary axis P. Alternatively or additionally, in examples in which the apparatus 1 includes an optical element (such as a lens assembly 3) with an optical axis, or an imaging element (such as an imager sensor 4) with an imaging axis, the primary axis P may be parallel to such an axis and / or may coincide with such an axis when the movable part 20 is in a central position or orientation (for example, see Figure 3A). In general, the movable part 20 may be movable relative to the support structure 10 with up to six degrees of freedom (DOFs). In the context of describing the DOFs of movement, the primary axis P may also be referred to as the z axis, and two further axes that are perpendicular to the primary axis P and to each other may be referred to as the x and y axes. The movable part 20 may be movable relative to the support structure 10 in all or in any subset (including only one) of the following DOFs: • Tx and Ty: Translational movement in the x-y plane. In other words, the movable part 20 may be independently movable along the x and y axes. The movable part 20 may be movable to any position in the x-y plane within a range of movement. Instead of such planar movement, the movable part 20 may be movable linearly, e.g. along the x or y axis. • Rx and Ry: Rotational movement (or simply rotation or tilting) about the x and y axes. In other words, the movable part 20 may be rotated about any line perpendicular to the primary axis P. The movable part 20 may be rotatable to any rotational position (i.e. to any orientation) within a range of movement. Instead of such two-axis rotation, the movable part 20 may be rotatable about a single axis, e.g. about the x or y axis. • Tz: Translational movement along the z axis. The movable part 20 may be movable to any translational position along the z axis within a range of movement. • Rz: Rotational movement (or simply rotation) about the z axis. The movable part 20 may be rotatable to any rotational position (i.e. to any orientation) within a range of movement. In some examples, the movable part 20 may be supported, e.g. by the bearing arrangement 40, so as to allow translational movement in the x-y plane (Tx, Ty) and / or rotational movement about the z axis (Rz). Translational movement along the z axis (Tz) and rotational movement about the x and y axes (Rx, Ry) may be constrained. Such support may be provided, for example, with a bearing arrangement 40 with a suitable arrangement of ball bearings or plain bearings which produce bearing forces in the +z direction and a biasing arrangement which produces a biasing force in the -z direction. Examples of actuator assemblies with such a bearing arrangement are disclosed in WO 2013 / 175197 Al and WO 2017 / 072525 Al, each of which is herein incorporated by reference. In some examples, the movable part 20 may be supported so as to allow tilting about the x and y axes (Rx, Ry) and optionally rotation about the z axis (Rz). The other DOFs of movement (i.e. Tx, Ty, Tz, Rz, or Tx, Ty, Tz) may be constrained. Such support may be provided by the bearing arrangement 40, for example in the form of a gimbal. Examples of such a bearing arrangement 40 are disclosed in WO 2021 / 209770 Al, which is herein incorporated by reference. Alternatively, such support may be provided exclusively by the actuating units 30, similarly to WO 2011 / 104518 Al which discloses an actuator assembly with 8 SMA wires connected between the support structure 10 and the movable part 20. WO 2011 / 104518 Al is herein incorporated by reference. In some examples, the movable part 20 may be supported so as to allow three-dimensional translational movement (Tx, Ty, Tz), while rotational movement (Rx, Ry, Rz) may be constrained. Such support may be provided by the bearing arrangement 40, for example in the form of nested linear bearings. Examples of such a bearing arrangement 40 are disclosed in WO 2021 / 209769 Al, which is herein incorporated by reference. Alternatively, such support may be provided exclusively by the actuating units 30, similarly to WO 2011 / 104518 Al. The movable part 20 may, alternatively or additionally, move in other DOFs. The movable part 20 may move in DOFs that are a combination of any two or more of Tx, Ty, Tx, Rx, Ry and Rz. For example, the movable part 20 may move along a helical path (i.e. move helically) about the z axis, and so concurrently move along the z axis and rotate about the z axis. In other words, Tz and Rz movement may be coupled. An example of such a helical actuator assembly is disclosed in WO 2019 / 243849 Al, which is herein incorporated by reference. The actuating units 30 are connected between the support structure 10 and the movable part 20. The actuating units 30 are arranged to apply actuating forces F (see e.g. Figs. 4 and 5) between the movable part 20 and the support structure 10. Selectively varying the actuating forces F may cause the movable part 20 to move relative to the support structure 10, for example within the DOFs allowed by the bearing arrangement 40. The actuating units 30 are thus capable of driving movement of the movable part 20 relative to the support structure 10. The bearing arrangement 40 may cause the movable part 20 to move in directions which differ from the directions of the actuating forces F. In simple examples of this, one component of each actuating force F causes the movement of the movable part 20, and another component of each actuating force F acts against the forces produced by the bearing arrangement 40. The camera module 1 also includes a lens assembly 3 and an image sensor 4. The lens assembly 3 includes one or more lenses configured to focus an image on the image sensor 4. The lens assembly 3 defines an optical axis O. The lens assembly 3 may include a lens carrier, for example in the form of a cylindrical body, supporting the one or more lenses. The image sensor 4 captures an image and may be of any suitable type, for example a charge coupled device (CCD) or a complementary metal-oxide-semiconductor (CMOS) device. The camera module 1 may be a compact camera module in which each lens has a diameter of 20mm or less, for example of 12mm or less. In the ("sensor-shift") variation of the camera module 1 shown in Figure 1A, the movable part 20 includes the image sensor 4. The lens assembly 3 may be fixed relative to the support structure 10, or may be movable relative to the support structure 10 along the optical axis O, as described below. In the ("lens-shift") variation, the image sensor 4 is fixed relative to the support structure 10 and the movable part 20 includes the lens assembly 3. The lens assembly 3 may be movable relative to the movable part 20 along the optical axis 0, as described below. In both of these variations, the actuator assembly 2 is configured to move the lens assembly 3 relative to the image sensor 4 in any direction in the plane perpendicular to the primary axis P and hence the optical axis O. Such movement has the effect of moving the image on the image sensor 4 and enables optical image stabilisation (OIS) to be implemented in the camera module 1. In the sensor-shift variation, the movable part 20 may also be rotatable about the primary axis P so as to also enable compensation for roll. In the ("module-tilt") variation shown in Figure IC, the movable part 20 includes both the lens assembly 3 and the image sensor 4. Again, the lens assembly 3 may be movable relative to the movable part 20 along the optical axis O, as described below. The actuator assembly 2 is configured to tilt the movable part 20 about two axes perpendicular to the primary axis P and to each other, and optionally rotate the movable part 20 about the primary axis P, enabling OIS to be implemented in the camera module 1. In the ("autofocus") variation shown in Figure ID, the movable part 20 includes the lens assembly 3, and the actuator assembly 2 moves the movable part 20 relative to the support structure 10 along the primary axis P and hence the optical axis O. Such movement has the effect of adjusting the focus of the image on the image sensor 4. So, auto-focus (AF) or zoom functionality can be implemented in the camera module 1. In some examples (not shown), the camera module 1 may include a first actuator assembly for providing OIS as illustrated in Figures 1A-C, and a second actuator assembly for providing AF as illustrated in Figure ID. One or both of the first and second actuator assemblies may correspond to actuator assemblies 2 as described herein. One of the first and second actuator assemblies may be another type of SMA actuator assembly or may be a non-SMA actuator assembly, e.g. a voice-coil motor (VCM) actuator assembly. As will be appreciated, in the lens-shift and module-tilt variations, the support structure 10 of the second actuator assembly 2 is fixed (or corresponds) to the movable part 20 of the first actuator assembly 2. In the ("AF+OIS") variation shown in Figure IE, the movable part 20 includes the lens assembly 3 and the actuator assembly 2 produces three-dimensional translational movement of the movable part 20 relative to the support structure 10, enabling both AF and OIS to be implemented using one actuator assembly 2. Other variations are also possible. For example, in the autofocus variation or the AF+OIS variation, the movable part 20 may include the image sensor 4 rather than the lens assembly 3. The camera module 1 may include combinations of the above described features, e.g. (a) lens shift and sensor shift, (b) module tilt and lens shift or sensor shift and autofocus, or (c) module tilt and AF+OIS. The camera module 1 also includes a controller 8. The controller 8 may be implemented in an integrated circuit (IC) chip. The controller 8 generates drive signals for the actuating units 30, in particular for SMA wires 34 forming part of the actuating units 30. SMA material has the property that on heating it undergoes a solid-state phase change that causes the SMA material to contract. Thus, applying drive signals to the SMA wires 34, thereby heating the SMA wires 34 by allowing an electric current to flow, will cause the SMA wires 34 to contract and thus actuate the actuating unit 30 so as to move the movable part 20. The drive signals are chosen to drive movement of the movable part 20 in a desired manner, for example so as to achieve OIS by stabilizing the image sensed by the image sensor 4 or to achieve AF / zoom by adjusting the focus of the image sensed by the image sensor 4. The controller 8 supplies the generated drive signals to the SMA wires 34. Optionally, the camera module 1 also includes a motion sensor (not shown), which may include a 3-axis gyroscope and a 3-axis accelerometer. The motion sensor can generate signals representative of the motion (specifically vibrations or "shake") of the camera module 1, which can be processed so as to produce signals representative of the required movement of the movable part 20 to compensate for such shake. The controller 8 receives such signals and can generate the drive signals for the SMA wires 34 to achieve OIS. Although the actuator assembly 2 is described in connection with a camera module 1, it will be appreciated that the actuator assembly 2 may be used in any device in which movement of a movable part 20 relative to a support structure 10 is desired, e.g. to provide haptic feedback in a haptic feedback device or to move a projector or display in an augmented reality (AR) or virtual reality (VR) device. Actuating unit Figure 3A shows a perspective view of an example of the actuating unit 30. Figure 3B shows part of the actuating unit 30 in plan view. A single actuating unit 30 is shown in Figures 3A and 3B, but it will be appreciated that the actuator assembly 2 generally has multiple actuating units 30, each of which may include the same components described with reference to Figures 3A and 3B. The actuating unit 30 includes a body portion 31 to which several other components of the actuating unit 30 are connected as described below. Typically, the body portion 31 is relatively rigid compared to the other components of the actuating unit, and does not deform significantly on actuation of the actuating unit 30. In some examples, the body portion 31 is not a distinct part of the actuating unit 30. For example, the body portion 31 may be defined as part of one of the other components of the actuating unit 30 or simply as a connection point between other components of the actuating unit 30. The actuating unit 30 also includes a force-modifying flexure 32. The force-modifying flexure 32 is connected between the body portion 31 and the support structure 10. One end of the force-modifying flexure 32 is connected to the body portion 31. The other end of the force-modifying flexure 32 is connected to the support structure 10, e.g. via a foot portion 36. The foot portion 36 is fixed relative to the support structure 10. In the depicted design, the force-modifying flexure is formed integrally with the foot portion 36 and with the body portion 31, for example from a single sheet of material (such as metal). The force-modifying flexure 32 allows the body portion 31 to pivot relative to the support structure 10 about an effective pivot point P. Although the effective pivot point P is shown in Figure 3B as being positioned in the middle of force-modifying flexure 32, the effective pivot point P may have a different position and also need not lie on the force-modifying flexure 32. Such pivotal movement of the body portion 31 relative to the support structure 10 is initially in a direction that is substantially perpendicular to the force-modifying flexure 32. The actuating unit 30 also includes an SMA element 34. In this example, the SMA element 34 is an SMA wire 34. The SMA wire 34 is connected between the body portion 31 and the support structure 10. One end of the SMA wire 34 is connected to the support structure 10, e.g. by a crimp 15. The other end of the SMA wire 34 is connected to the body portion 31, e.g. by a crimp 35. The actuating unit 30 also includes a coupling link 33. In this example, the coupling link 33 is a coupling flexure 33. The coupling flexure 33 is connected between the body portion 31 and the movable part 20. One end of the coupling flexure 33 is connected to the body portion 31. The other end of the coupling flexure 33 is connected to the movable part 20. The coupling link 33 transfers or transmits an actuating force F from the body portion 31 to the movable part 20. The coupling link 33 is compliant (i.e. deformable) in a direction (or in multiple directions) perpendicular to the actuating force F. This allows the movable part 20 to move in directions other than the direction of the coupling flexure 33 and actuating force F. This can be needed, for example, where different actuating units 30 cause the movable part 20 to move in different directions. The SMA wire 34 is arranged, on contraction, to apply an input force Fi on the body portion 31. The input force Fi acts parallel to the length of the SMA wire 34. The force-modifying flexure 32 and the body portion 31 are arranged to modify the input force Fi so as to give rise to the actuating force F, which is transmitted from the body portion 31 to the movable part 20 by the coupling flexure 33. In particular, the input force Fi deforms the force-modifying flexure 32, thereby causing the body portion 31 to pivot about the effective pivot point P. In simple terms, the force-modifying flexure 32 and the body portion 31 act like a lever. The force-modifying flexure 32 and the body portion 31 may modify the direction and / or the magnitude of the input force Fi so as to give rise to the actuating force F. In the example illustrated in Figures 3A and 3B, the coupling flexure 33 is at an angle of '90° relative to the SMA wire 34. Also, in this example, the force-modifying flexure 32 is arranged at an angle a of -30° relative to the SMA wire 34, and the force-modifying flexure 32 is placed in tension on contraction of the SMA wire 34. Hence, on contraction of the SMA wire 34 and on resulting deformation of the forcemodifying flexure 32, the body portion 31 initially moves at an angle of-60° (90°-a) relative to the length of the SMA wire 34. Thus, it will be appreciated that, in this example, the force is de-amplified and the stroke is amplified, while the direction of the forces / movements is changed by an angle of -90°. More generally, the change in direction of the force depends on the angle between the SMA wire 34 and the coupling flexure 33. Also more generally, the change in magnitude of the force is dependent on the ratio of i) the distance Ds from the effective pivot point P to the line on which the SMA wire 34 lies and ii) the distance De from the effective pivot point P to the line on which the coupling flexure 33 lies. In particular, F / Fi is proportional to Ds / Dc. If the SMA wire 34 lies on a line that is closer to the effective pivot point P than the line on which the coupling flexure 33 lies, then the input force Fi is de-amplified. At the same time, the movement of the movable part 20 is amplified, i.e. increased relative to a change in length of the SMA wire 34. Alternatively, if the SMA wire 34 lies on a line that is further away from the effective pivot point P than the line on which the coupling flexure 33 lies, then the input force Fi is amplified. At the same time, the movement of the movable part 20 is de-amplified, i.e. decreased relative to a change in length of the SMA wire 34. The actuating unit 30 can thus be configured to amplify movement or to amplify force due to contraction of the SMA wire 34. The actuating unit 30 can also be configured to change the direction of the input force Fi. In some examples, the actuating unit 30 is configured to change the direction of the input force Fi without changing the magnitude of the force or movement. The ratio Ds / Dc is dependent on the location of the end of the SMA wire 34 that is connected to the body portion 31 and the location of the end of the coupling flexure 33 that is connected to the body portion 31. By way of example, the distance Ds could be increased by connecting the coupling flexure further to the left of body portion 31 shown in Figure 3B, thereby decreasing Ds / Dc and so increasing the amount of stroke amplification. The ratio Ds / Dc is also dependent on the orientation of the SMA wire 34 and the orientation of the coupling flexure 33. Such orientations can be defined with reference to the forcemodifying flexure 32 (as above) or any suitable reference line. By way of example, the distance Ds could be decreased by angling the SMA wire 34 shown in Figure 3B so that it passes closer to the effective pivot point P, thereby decreasing Ds / Dc and so increasing the amount of stroke amplification. In summary, the amount by which the force-modifying flexure 32 amplifies or de-amplifies the force / stroke of the SMA wire 34 may be tailored by: • adjusting the angle of SMA wire 34 (and thus the input force Fi); • adjusting the location of the connection point between the SMA wire 34 and the body portion 31 (and thus the location at which the input force Fi acts on the body portion 31); • adjusting the angle of the coupling flexure 33 (and thus the actuating force F); and / or • adjusting the location of the connection point between the coupling flexure 33 and the body portion 31 (and thus the location from which the body portion 31 applies the actuating force F). In some examples, at least one actuating unit 30 (preferably each actuating unit 30) is configured such that the force-modifying flexure 32 amplifies an amount of contraction of the SMA wire 34. Such amplification, for example, may be by a factor greater than 1.5, preferably greater than 2, further preferably greater than 3. For this purpose, in the example illustrated in Figures 3A and 3B, the angle a between the SMA wire 34 and the force-modifying flexure 32 may be in the range from 0 to 45 degrees, preferably from 13 to 40 degrees. However, in general, the angle a may have other values and the connection points of the SMA wire 34 and / or coupling flexure 33 to the body portion 31 may be adjusted to achieve a desired amount of amplification. As described above, in the example illustrated in Figures 3A and 3B, the coupling flexure 33 is at an angle of about 90 degrees relative to the SMA wire 34. This allows the actuating unit 30 to fold around a corner of the movable part 20 in a compact manner. The angle between the coupling flexure 33 and the SMA wire 34 may be in the range from 70 to 110 degrees, preferably from 80 to 100 degrees. However, in general, the angle between coupling flexure 33 and SMA wire 34 may be outside these ranges. For instance, in the actuating unit 30 illustrated in Figure 3C, the coupling flexure 33 is substantially perpendicular to the SMA wire 34. In the above-described examples, the actuating unit 30 is arranged in a plane. In particular, the SMA wire 34, the coupling flexure 33 and the force-modifying flexure 32 are arranged so as to substantially extend in a common plane, at least when the actuator assembly 2 is in an initial configuration. This allows for a compact configuration of the actuating unit 30. The body portion 31, when embodied by a plate, may further be arranged to extend in the plane. However, in general, the components of the actuating unit 30 need not be arranged in a common plane. The SMA wire 34 and / or the coupling flexure 33 may be angled relative to the plane, for example. In the above-described examples, the force-modifying flexure 32 is placed in tension on contraction of the SMA wire 34. This reduces the risk of buckling of the force-modifying flexure 32, reducing the risk of damage to the actuator assembly and making the actuator assembly 2 more reliable. However, the forcemodifying flexure 32 could instead be arranged so as to be placed under compression on contraction of the SMA wire 34. With reference to Figure 3B, for example, the force-modifying flexure 32 could extend to the bottom-right from the connection point between the body portion 31 and the force-modifying flexure 32, and so be placed under compression on contraction of the SMA wire 34. An arrangement in which the force-modifying flexure 32 is placed under compression is disclosed in WO 2022 / 084699 Al, which is herein incorporated by reference. In the above-described examples, the force-modifying flexure 32 and the SMA wire 34 connect at one end to the support structure 10, and the coupling flexure 33 connects at one end to the movable part 20. In general, this arrangement may also be reversed, with the force-modifying flexure 32 and the SMA wire 34 connecting at one end to the movable part 20, and the coupling flexure 33 connecting at one end to the support structure 10. In the above-described examples, the actuating unit 30 includes a coupling link 33 in the form of a coupling flexure 33. The purpose of the coupling link 33 is to allow movement of the movable part 20 in directions perpendicular to the actuating force F. In general, however, the actuating unit 33 need not include a coupling link 33, for example in examples in which there is no movement of the movable part 20 in directions perpendicular to the actuating force F. Furthermore, the coupling link 33 may be embodied by components other than the coupling flexure 33, for example by a ball bearing or plain bearing configured to transmit the actuating force F to the movable part 20 while allowing movement of the movable part 20 in directions perpendicular to the actuating force F. Such alternative examples of the coupling link 33 are disclosed in WO 2022 / 084699 Al. Arrangement of four actuating units Figure 4 schematically shows a plan view of an example of the actuator assembly 2, showing an arrangement of actuating units 30. In this example, the actuator assembly 2 includes a total of four actuating units 30. The four actuating units 30 may apply actuating forces F between the movable part 20 and the support structure 10. The actuating forces F are applied to the movable part 20 relative to the support structure 10. The arrangement of actuating units 30 of Figure 4 may be used, for example, in examples in which the movable part 20 is movable relative to the support structure 10 in a movement plane. So, Tx, Ty and optionally Rz movement of the movable part 20 may be allowed. The four actuating units 30 of Figure 4 are in an arrangement capable of applying actuating forces F so as to move the movable part 20 relative to the support structure 10 to any positions within a range of movement. The range of movement may be within a movement plane that is perpendicular to the primary axis P. In particular, two actuating units 30 (e.g. the top and bottom actuating units in Figure 4) are arranged to apply actuating forces F in opposite directions parallel to a first axis (e.g. the x axis in Figure 4). The other two of actuating units (e.g. the left and right actuating units in Figure 4) are arranged to apply actuating forces F in opposite directions parallel to a second axis (e.g. the y axis in Figure 4), perpendicular to the first axis. By appropriately varying the difference in actuation amount between the opposing actuating units 30, the movable part 20 may thus be moved independently along the first and second axes. The opposing actuating forces F are not colinear, but offset from each other in a direction perpendicular to the actuating forces. Providing opposing actuating units 30 allows the tension in the SMA wires 30 of the respective actuating units 30 to be controlled, allowing for more accurate and reliable positioning of the movable part 20 compared to a situation in which actuating units 30 do not oppose each other. In examples, none of the actuating forces F are collinear. This allows the arrangement of actuating units 30 to translationally move the movable part 20 without applying any net torque to the movable part 20. So, the movable part 20 can be moved translationally in the movement plane without rotating the movable part 20 in the movement plane. In general, the arrangement of actuating units 30 is capable of accurately controlling a torque or moment of the movable part 20 about the primary axis P. So, the actuating units 30 are capable of rotating (or not rotating) the movable part 20 relative to the support structure about the primary axis P. In particular, two actuating units 30 (e.g. the top and bottom actuating units in Figure 4) are arranged to apply actuating forces F so as to generate a torque or moment between the movable part 20 and the support structure 2 in a first sense (e.g. clockwise) around the primary axis P. The other two actuating units 30 (e.g. the left and right actuating units 30 in Figure 4) are arranged to apply actuating forces F so as to generate a torque or moment between the movable part 20 and the support structure 2 in a second, opposite sense (e.g. anti-clockwise) around the primary axis P. This allows the movable part 20 to be rotated by simultaneously increasing or decreasing the tension of SMA wires in any of the two actuating units 30. As shown, two actuating units 30 may be arranged to apply actuating forces in a corner of the actuator assembly 2. The other two actuating units 30 may be arranged to apply actuating forces in another, opposite corner of the actuator assembly 2. The actuator assembly 2, and in particular the movable part 20 and / or the support structure 10, may have a square or rectangular footprint. Each actuating unit 30 may be provided on one of the four sides of the actuator assembly 2. In particular, each actuating unit 30 may bend around a corner of the movable part 20 such that the SMA wire 34 and the coupling flexure 33 of each actuating unit 30 extend along adjacent edges of the movable part 20. So, the actuating unit 30 may be as configured in Figures 3A and 3B, for example. The four SMA wires 32 of the four actuating units 32 may extend along the four different edges of the movable part 20. The arrangement of actuating forces F applied between movable part 20 and support structure 10 corresponds to the arrangement of SMA wires 30 described in WO2013 / 175197 Al, which is herein incorporated by reference. In this example, the actuating forces F are perpendicular to the primary axis P, and may be parallel to the movement plane. However, in general the actuating forces F may be angled relative to the movement plane. The actuating forces F may thus have a component along the primary axis P. This component along the primary axis P may be resisted by the bearing arrangement 40, for example, to provide movement of the movable part 20 in degrees of freedom allowed by the bearing arrangement 40. In some examples, it may even be desirable for actuating forces F to have a component in parallel to the primary axis P, for example so as to load plain or rolling bearings arranged between the movable part 20 and the support structure 10. Although, for illustrative purposes, the arrangement of actuating units 30 was described as moving the movable part 20 in the movement plane (e.g. translationally along the x and y axis, or rotationally about the primary axis P), in other examples the movable part 20 may be moved differently. For example, the same arrangement of actuating forces F may be used to tilt the movable part 20 relative to the support structure 10 about axes perpendicular to the primary axis, due to appropriate movement constraints provided by the bearing arrangement 40. For example, the bearing arrangement 40 may include a plurality of flexures for guiding tilting of the movable part 20 about the axes perpendicular to the primary axis P. Examples of such bearing arrangement 40 are described in WO2022 / 029441 Al, which is herein incorporated by reference. Although the actuator assembly 2 is described herein in the context of four actuating units 30, in general the actuator assembly 2 may include fewer actuating units 30. For example, the actuator assembly 2 may include two actuating units 30, e.g. the two actuating units 30 depicted in the top left of Figure 4. The forces applied to the movable part 20 by the two actuating units 30 may be opposed by a biasing force of one or more resilient elements, such as springs. With reference to Figure 4, the two actuating units 30 in the bottom right corner may be replaced with springs applying biasing forces along the corresponding depicted arrows, for example. Arrangement of eight actuating units Figures 5 schematically shows a perspective view of an actuator assembly 2 with a total of eight actuating units 30. The eight actuating units 30 may apply actuating forces F between the movable part 20 and the support structure 10. The actuating forces F are applied to the movable part 20 relative to the support structure 10. The arrangement of actuating units 30 of Figures 5A and 5B may be used, for example, in examples in which the movable part 20 is movable relative to the support structure 10 in three translational degrees of freedom (Tx, Ty, Tz) (see Figure IE) or in two or three rotational degrees of freedom (Rx, Ry or Rx, Ry, Rz) (see Figure IC). The eight actuating units 30 may be arranged such that the actuating forces F of the eight actuating units 30 are oriented or arranged in a manner equivalent to the orientation or arrangement of the forces applied by the eight SMA wires in the actuator assemblies disclosed in WO 2011 / 104518 Al. More specifically, the actuating forces F (e.g. when visualised as vectors at particular positions in space) are arranged on each of four sides (i.e. a first side, a second side, a third side and then a fourth side) around the primary axis P. The two actuating forces F on each side are inclined in opposite senses with respect to each other, as viewed perpendicular from the primary axis. The four sides on which the actuating forces F are arranged extend in a loop around the primary axis. In this example, the sides are perpendicular and so form a square as viewed along the primary axis, but alternatively the sides could take a different e.g. quadrilateral shape. In this example, the actuating forces F are parallel to the outer faces of the square envelope of the moveable part 6 but this is not essential. Four actuating forces F, including one force on each of the sides, form a 'first' group that have a component in one direction ('upwards' or +z) and the other four actuating forces F form a 'second' group that have a component in the opposite direction ('downwards' or -z). Herein, 'up' and 'down' refer to opposite directions along the primary axis P. The actuating forces F have a symmetrical arrangement in which their magnitudes and inclination angles are the same, so that both the first group of actuating forces F and the second group of actuating forces F are each arranged with two-fold rotational symmetry about the primary axis. As a result of this symmetrical arrangement, different combinations of the actuating forces F are capable of driving movement of the moveable part 6 with multiple degrees of freedom, as follows. The first group of actuating forces F, when generated together, drive upwards (+z) movement, and the second group of actuating forces F, when generated equally, drive downwards (-z) movement. Within each group, adjacent pairs of actuating forces F, when differentially generated, drive tilting about a lateral axis perpendicular to the primary axis z (Rx or Ry). Tilting in any arbitrary direction may be achieved as a linear combination of tilts about the two lateral axes. Sets of four actuating forces F, including two actuating forces F from each group, when generated together, drive movement along a lateral axis perpendicular to the primary axis P (Tx orTy). Movement in any arbitrary direction perpendicular to the primary axis z may be achieved as a linear combination of movements along the two lateral axes. The SMA actuator may have other specific arrangements of actuating units 30 to those shown in Figure 5. For example, strict symmetry is not required. Furthermore, instead there being an up-pulling actuating unit 30 and a down-pulling actuating unit 30 on each side, there maybe two up-pulling actuating units 30 on each of two opposite sides (e.g. the first and third sides) and two down-pulling actuating units 30 on the other two sides (e.g. the second and fourth sides). Arrangement of actuating units In general, the present invention relates to particularly advantageous arrangements of two or more actuating units 30. Specifically in miniature applications, such as cameras for smartphones or other applications in handheld devices, the space available for the actuator assembly 2 may be constrained. The present invention is concerned with maximising performance of the actuator assembly 2 within a given space constraint and / or making the actuator assembly 2 more compact. The arrangement of actuating units 2 described below, in particular the manner in which plural actuating units 2 overlap with each other and / or the first and second parts, is particularly advantageous when applied the arrangement of four actuating units 30 described in connection with Figure 4. The plural actuating units 30 may generally extend in parallel planes. The actuating planes within which the plural actuating units30 actuate (i.e. within which the body portion 31 of the plural actuating units 30 may move on actuation) may coincide or be parallel to each other. Overlapping coupling flexures for reduced height The inventors have found that a particularly advantageous arrangement of two actuating units 30 is to allow the coupling flexures 33 of the two different actuating units 30 to overlap. So, according to some embodiments, the coupling flexures 33 of two different actuating units 30 (in particular of adjacent actuator units 30) overlap when viewed along the primary axis P. The two different actuating units 30 are also referred to as a pair of actuating units 30 herein. In particular, only the coupling flexures 33 of the two actuator units 30 overlap. The body portions 31 and force-modifying flexures 32 of adjacent actuator units 30 do not overlap when viewed along the primary axis P. This is apparent from Figure 6B, for example. One of the two different actuating units 30 applies a respective actuating force F parallel along the x axis, and the other of the two different actuating units 30 applies a respective actuating force F parallel to the y axis. The coupling flexures 33 of the two actuating units 30 not shown in the cut-out of Figure 6B also overlap when viewed along the primary axis P. In particular, the coupling flexures 33 of the two different actuating units 30 cross over or intersect when viewed along the primary axis. The coupling flexures 33 of the two different actuating units 30 may form an angle of about 90 degrees therebetween, preferably an angle in the range from 70 to 110 degrees, preferable from 80 to 100 degrees. Providing such overlap between coupling flexures 33 allows the height of the actuator assembly 2 to be reduced while maintaining a relatively compact footprint, compared to a situation in which the coupling flexures 33 do not overlap. This is apparent, for example, from a comparison with the actuator assembly 2 depicted in Figures 8A and 8B. In the actuator assembly 2 of Figures 8A and 8B, the coupling flexures 33 do not overlap. Instead, overlap is provided between the body portions 31 of two adjacent actuating units 30. As a result, the height of the actuator assembly 2 of Figures 8A and 8B is greater than that of the actuator assembly 2 of Figures 6 and 7A-B. Furthermore, a hypothetical example in which neither the coupling flexures 33 nor other parts of the actuating units 30 of adjacent actuator assemblies 2 overlap would have a much greater footprint (in the x-y plane) than the actuator assembly 2 of Figures 6 and 7A-B. Allowing specifically the coupling flexures 33 of adjacent actuator units 30 to overlap thus provides a generally compact actuator assembly 2. Put another way, the body portion 31, the force-modifying flexure 32 and the foot portion 36 of one actuating unit 30 are located between the SMA wire of the one actuating unit 30 and the coupling flexure 33 of the other actuating unit 30 of the two adjacent actuating units 33. The distance between the SMA wire 34 of one actuating unit 30 and the coupling flexure 33 of the other actuating unit 30 is increased compared to the Figure 8 actuator assembly 2. Despite the overlap when viewed along the primary axis P, the coupling flexures 33 may be configured not to be in direct contact with each other upon movement of the movable part 20, in particular over the entire movement range of the movable part 20 relative to the support structure 10. This may be achieved, for example, by providing at least one coupling flexure 33 (optionally both coupling flexures 33) with a kinked portion (not shown) in the region of overlap. This kinked portion may also be referred to as a jog or as an offset, along the primary axis, formed in the coupling flexure 33. A major portion of both of the coupling flexures 33 may thus extend in the same plane, with only the kinked portion offset from the plane to avoid clashes between the coupling flexures 33. Alternatively, the coupling flexures 33 of the two actuating units 33 may be arranged to be offset along the primary axis P, thereby avoiding direct contact. Figure 9, for example, depicts an embodiment of an actuating units 30 in a multi-layer arrangement. In particular, the body portion 31 comprises at least two layers, stacked or offset from each other along the primary axis P. In the depicted embodiment, the force-modifying flexure 34 and the foot portion 36 also comprise two layers stacked along the primary axis P, but in general only one layer may also be provided. The coupling flexure 33 is a single flexure, with a height along the primary axis P that is less (in particular half) of the height of the body portion 31. So, in general, the coupling flexure 33 may have an extent along the primary axis P that is less than that of the body portion 31, in particular less than half of the extent of the body portion 31 along the primary axis P. This allows one of the two adjacent actuating units 30 to be arranged with one side up, and the other of two adjacent actuating units 30 to be arranged with the other side up. The two actuating units 30 may thus be identical in structure, but arranged on the actuator assembly 2 with different sides facing in a given direction along the primary axis P. Clashing of the coupling flexures 33 can thus be avoided. Figure 11B shows an embodiment with such an arrangement of actuator units 30. In particular, each actuating unit 30 may comprise a first surface parallel to a plane orthogonal to the primary axis P and a second surface parallel to the plane. The first and second surfaces are provided on opposite sides of the actuating unit 30. The two adjacent actuating units 30 may be arranged on the actuator assembly 2 with the first surface of one and the second surface of the other facing in the same direction along the primary axis P. As depicted in the embodiments of Figures 6,10 and 11, the SMA wire 34 extends, viewed orthogonally to the SMA wire, past a point at which the coupling flexure 33 connects to the body portion 31. In Figure 6B, for example, the SMA wire 34 extends further to the left than the connection point between coupling flexure 33 and body portion 31. As a result, the length of the SMA wire 34 is greater compared to a situation in which the SMA wire 34 stops at a point where the coupling flexure 33 connects to the body portion 31 (as in Figure 8B, for example). A longer SMA wire 34 allows for increased stroke and / or more accurate positioning control of the movable part 20. Figure 10 shows another embodiment of an actuating unit 30. Compared to the actuating unit 30 of Figures 6 and 8, the SMA wire 34 is arranged on the same side of the body portion 31 as a point at which the force-modifying flexure 32 connects to support structure 10. In particular, the foot portion 36 is provided on the same side of the body portion 31 as the SMA wire 34. This reduces the extent of the actuating unit 30 in a lateral direction, i.e. in a direction along the actuating force F. The footprint of the actuator assembly 2 may thus be reduced. In some embodiments, each coupling flexure 33 may be broader in a region of overlap with the other coupling flexure 33. So, in the region where the coupling flexures 33 overlap may be wider (in a direction perpendicular to the length of the coupling flexure) when viewed along the primary axis P than adjacent portions of the coupling flexure 33. This may reduce the risk of damage to the coupling flexures 33 in the region of overlap due to clashes of the coupling flexures, for example during impact events such as drops. Figure 10 schematically show such broader portions 33b of the coupling flexures 33. Figure 11A and 11B show further embodiments of an actuating unit 30. Compared to the actuating unit 30 of Figures 6 and 8, the SMA wire 34 is arranged so as to overlap with the body portion 31 when viewed along the primary axis P. This allows the extent of the actuating unit 30 in a lateral direction, i.e. in a direction along the actuating force F to be reduced. The footprint of the actuator assembly 2 may thus be reduced. Connecting actuating units to different parts As already discussed above, the coupling flexures 33 may connect to either one of the support structure 10 and movable part 20, and the SMA wires 34 and force-amplifying flexures 32 may respectively connect to either of the movable part 20 and support structure 10. Figures 12A and 12B show embodiments of the actuator assembly 2 in which some actuating units 30 are arranged such that their coupling flexures 33 connect to the movable part 20 and some other actuating units 30 are arranged such that their coupling flexures connect to the support structure 10. Figure 12A shows this conceptionally, whereas Figure 12B shows further details of the arrangement of actuator units 30. In particular, the actuating units 30 applying actuating forces F along the y axis are referred to herein as static actuating units 30s. These actuating units 30s are connected between the movable part 20 and the support structure 10 with the force-modifying flexure 32s and the SMA wire 34s connecting to the support structure 10, and the coupling flexure 33s connecting to the movable part 20. The actuating units 30 applying actuating forces F along the x axis are referred to herein as movable actuating units 30m. These actuating units 30m are connected between the movable part 20 and the support structure 10 with the force-modifying flexure 32m and the SMA wire 34m connecting to the movable part 20, and the coupling flexure 33m connecting to the support structure 10. Of course, in other arrangements, the static actuating units 30s may apply actuating forces F along the x axis and the movable actuating units 30m may apply actuating forces F along the y axis. In some other embodiments, one static actuating unit 30s may apply an actuating force F along the x axis and the other static actuating unit may apply an actuating force F along the y axis. Similarly, one movable actuating unit 30m may apply an actuating force F along the x axis (opposing the actuating force F of the corresponding static actuating unit 30s) and the other movable actuating unit 30m may apply an actuating force F along the y axis (opposing the actuating force F of the corresponding static actuating unit 30s). Providing both static actuating units 30s and movable actuating units 30m in combination may reduce the amount of overlap, when viewed along the primary axis, of the actuating units 30m. Indeed, as shown in Figures 12A and 12B, overlap may be avoided entirely. The height of the actuator assembly 2 may thus be reduced. This is achieved without increasing the footprint of the actuator assembly 2 significantly. Further details of actuator assembly Figures 13A and 13B show further details of the actuator assembly 2. In particular, Figure 13A shows an example of the static crimp 15 mounted to the support structure 10. Figure 13B shows an example of the foot portion 36 of the actuating unit 30 mounted to the support structure 10. As shown in Figure 13A, the static crimp 15 may be elevated or offset from a surface of the support structure 10. In particular, kinks or bends may be provided in a portion connected to or integrally formed with the static crimp 15 so as to elevate the static crimp. Similarly, as shown in Figure 13B, the foot portion 36 may include a kink or bend. This allows the body portion 31, as well as the crimp 35 connected to the body portion 31 and the force-modifying flexure 32 to be elevated or offset from a surface of the support structure 10. Both crimps 15, 35 connecting to the SMA wire 34 may thus be elevated above the surface of the support structure 10, allowing alignment of the two crimps 15, 35 in a common plane. The SMA wire 34 may extend parallel to the surface of the support structure 10 and / or parallel to the actuating forces F and / or parallel to the movement plane. The force-modifying flexures 32 and / or body portions 31 of different actuating units 30 may also be provided at different offsets along the primary axis P in this manner, thus avoiding clashes between actuating units 30 even when there is overlap when viewed along the primary axis P. Figure 13C shows a raised feature 20r on the movable part 20. The coupling flexure 33 of an actuating unit 30 may connect to this raised feature 20r. The height of this raised feature 20r may be designed so as to allow the coupling flexure 33 to extend substantially parallel to the surface of the support structure 10 and / or parallel to the actuating forces F and / or parallel to the movement plane. Actuating units overlapping the movable part Another aspect of the present invention relates to allowing the actuating units 2 to overlap, when viewed along the primary axis, with both the movable part 20 and the support structure 10 Referring in particular to Figures 14 to 18, several actuator assemblies 2 according to further aspects of the present invention will now be described. Here, the actuating units 30, when viewed along a primary axis, overlap (or are located within) the region bounded by the movable part 20. Herein, the primary axis may be an axis perpendicular to the plane ('movement plane') in which the movable part 20 is movable. By overlapping the movable part 20 in this way, rather than being positioned around the movable part 20, the actuating units 30 can have various different (and advantageous) configurations, as will become apparent. For instance, the SMA wire 34 can be longer than the length of any side of the movable part 20 without increasing the footprint of the actuator assembly 2. This may improve the performance of the actuator assembly while meeting given size constraints. In some embodiments, the support structure 10 and the movable part 20 comprise elements of an imaging system, such as the image sensor and / or lens assembly. The elements of the imaging system comprised in the support structure 10 and the movable part 20 may form opposite ends of the imaging system. The elements may bound a region between them. An optical path may extend between the elements, and the region may define the optical path. It may be desirable to keep the region free from components which may otherwise obstruct the optical path through the imaging system. Optionally, the actuating units 30 are located at least partly outside of the region bounded by the elements of the imaging system along the direction of the primary axis. For example, the flexure arrangements 72 may be positioned behind the image sensor 6 (i.e. below the image sensor 6 in the orientation shown in Figure 1). The image sensor 6 may be between, in the direction of the primary axis, the actuating units 30 and the lens 22. "Dual-plane" actuator assembly Figure 14 is a schematic plan view of an actuator assembly 2, including four actuating units 30 in two planes. In particular, the actuating units 30 may be arranged in two parallel actuating planes. The actuating units 30 will be described with reference to flexure arrangements 72a-d. The flexure arrangements 72a-d are comprised by the actuating units 30. The flexure arrangements 72a-d correspond to the combination of body portion 31, force-modifying element 32 and optionally coupling link 33. The depicted flexure arrangements 72a-d are functionally similar to the combination of force-modifying flexure 32, coupling flexure 33 and body portion 31 described in relation to Figures 3A and 3B, apart from having a different configuration. This configuration uses both (i) the concept of shallow angle wire to bend the flexure fulcrum beam orthogonal to itself and (ii) the concept of a first-class lever to amplify the displacement, trading off force for stroke. The shallow angle refers to the angle between the SMA wire 34 and the force-modifying flexure 32. A shallow angle means an angle of less than 45°, optionally less than 40°, optionally less than 30°, optionally less than 20° and optionally less than 10°. For example, in one of the actuating units 30, the force-modifying flexure 32 extends from the upper-left of the actuator assembly 2 as illustrated, and the SMA wire extends towards the lower-right of the actuator assembly 2. For the angle wire amplification, the SMA wire 34 is attached to the moving crimp 35 and angled shallowly with the force-modifying flexure 32. This allows amplification orthogonal to the flexure direction as it must bend a lot to accommodate small changes in wire length. For the first-class lever amplification, the pivot fulcrum is provided by the force-modifying flexure 32 allowing the rotation of the body portion 31 (which may be shaped as a beam) at that location. The amplified displacement is then carried to the actuator via the coupling flexure 33 which allows the transfer of displacement but also take bending when the actuator moves in the opposite direction of what this individual flexure system's positive displacement is. The amplification factor can be adjusted by adjusting the ratio of (a) the distance between the wire load (i.e. the crimp 35 where the SMA wire 34 exerts a force on the body portion 31) and the fulcrum (i.e. the position where the force-modifying flexure 32 joins the body portion 31) to (b) the distance between the displacement amplified load (i.e. the position where the body portion 31 joins the coupling flexure 33) and the fulcrum. The relationship between this ratio and the amplification factor is about 1:1 but due to the flexure pivot bending, additional stroke will be observed. As shown in Figure 14, optionally, the angle between the coupling flexure 33 and the sides of the support structure 10 is in the range of 30 to 60 degrees. As shown in Figure 14, optionally, the angle between the force-modifying flexure 32 and the sides of the support structure 10 is in the range of 30 to 60 degrees. An advantage of having the coupling flexure 33 and force-modifying flexure 32 at ~45 degrees is that they can take bending when the other one would have been in compression. This reduces or eliminates the possibility of a buckle failure. The force-modifying flexure 32 and the coupling flexure 33 are both thin and highly flexible and in arranged in tension. In order to maximise wire length an additional section, i.e. the body portion 31, is created (e.g. by forming the body portion 31 as a beam) between the two which is in compression. As this section does not need to be flexible to either create high amplification or to couple the parts and allow orthogonal movement it can be very wide and stiff. This is advantageous to avoid reduction of stroke, and is also useful for helping the compression section, namely the body portion 31, avoid buckling. The compression section helps to make the package as small as possible and therefore fit behind an image sensor 6, for example, whilst still having sufficient wire length to achieve high stroke. Optionally the actuator assembly 2 has a tensile flexure design. This allows the positioning of the actuator completely behind the image sensor 6. This can reduce the footprint of the combination of the actuating units and the image sensor 6. This is due to the wire stroke increase allowing high stroke but in a small footprint, small enough to be behind the sensor. Wires for this tensile flexure amplification also have the advantage of routing the wire across the actuator to make full use of the footprint of the actuator for maximum length wire. Compared to the standard 4 wire, the maximum possible wire length increase from the length of the actuator side length tothecorner-to-corner length of the actuator. "Single plane" actuator assembly The actuating units 30 can be located in a single plane (as shown in Figure 15) or in multiple planes (as shown in Figure 14) to aid packaging or aid material properties. For the single-plane version, all the flexure systems can be etched together in a single sheet of etching, saving cost and complexity. Figure 15 is a schematic plan view of an actuator assembly 2, including four actuating units in one plane. Figure 16 is a schematic perspective view of the actuator assembly 2 illustrated in Figure 15. "Class 3 lever" actuator assembly Figure 17 is a schematic plan view of an actuator assembly 2 including four actuating units 30 with a different configuration. In this configuration, the amplification comes from a third-class lever action. The force-modifying flexure 32 allows the pivoting of the amplification beam (i.e. the part of the body portion 31 extending between the force-modifying flexure 32 and the coupling flexure 33) and input force beam (i.e. the part of the body portion 31 extending between the crimp 35 and the force-modifying flexure 32). Taking the flexure arrangement 72b shown in the top left of the diagram as an example, there is a shallow angle between the SMA wire 34 and the force-modifying flexure 32. The SMA wire 34, on contraction, applies a force that has a major component in the direction of the force-modifying flexure 32 (thereby keeping it in tension) and a minor component in the downward direction in the orientation shown in the diagram. This minor component causes the force-modifying flexure 32 and the body portion 31 to rotate clockwise. Again, the amplification factor can be adjusted by adjusting the ratio of the distance between the wire load and the fulcrum to the distance between the displacement amplified load and the fulcrum. This configuration includes coupling flexures 33 of different actuating units which are not orthogonal to one another. This results in a greater stroke in one of the directions than the other. In general, it may be preferable to have orthogonal coupling flexures 33 to have even stroke in each direction. Bearings for the actuating units Figure 18 is a schematic diagram of a force-modifying flexure 32 and an SMA wire 34 biasing the actuating unit against a bearing 55. The top right side of the diagram shows a schematic plan view of an actuating unit. The bottom left side of the diagram shows a side on view of the same actuating unit. The bottom left drawing is provided to show that the force-modifying flexure 32 (and / or the coupling flexure 33) and / or the SMA wire 34 may extend in the direction of the primary axis (i.e. extend between planes) so as to apply a force to keep the flexure arrangement 72 in contact with the bearing 55. The wire loading can be used to bias the flexures down on to the bearing system, by angling the wires in the optical axis slightly or by angling the flexures, also in the optical axis, so that under tension they pull the system on to the bearings 55. It is possible that the flexure arrangements 72 may not resist out of plane forces. As shown in Figure 18, optionally bearings 55 which prevent movement out of the plane (i.e. in the optical direction) are provided. One way of applying these bearings 55 is to use plain bearings under the wide compression section (i.e. the body portion 31) of the design, between this part and the support structure 10 and between the compression section of the flexure arrangement 72 and the movable part 20 (which may comprise an image sensor 6). An alternative option would be to use rolling bearings such as ball bearings in one or both of these gaps. Furthermore it is possible to preload the system in the optical direction so only one set of bearings 55 is provided. For example the force-modifying flexure 32 and / or the coupling flexure 33 could be preloaded towards the support structure 10 and then a plain bearing 55 may be placed between the support structure 10 and the flexure. Additionally by placing the bearing 55 on the compression section, it will reduce the potential for the compression section to buckle. Such bearings 55 can also be used for other flexure arrangements such as those included in the reference example (which, for example, do not overlap with movable part 20). Integrated crimps Figure 19 is a schematic diagram of an actuating unit with an integrated crimp 35. The crimp etching may be integrated on the flexure system etching to reduce or minimise the amount of etching required and to avoid the need to weld on crimps. Alternative configurations Some alternative configurations of actuator assemblies 2 are shown in Figures 20 to 24. Figure 20 is a schematic plan view of one such actuator assembly 2 with a particularly simple layout. Optionally, the flexure arrangements 72 are provided using a single etching, i.e. from a single sheet of metal. Figure 21 is a schematic plan views of another actuator assembly 2. Again, the flexure arrangements 72 may be provided using a single etching, i.e. from a single sheet of metal. This configuration may be particularly suitable for apparatuses, e.g. cameras that have a square footprint. As will be apparent, the stroke amplification may be particularly high. Figure 22 is a schematic plan view of another actuator assembly 2. This configuration is a reverse of the configuration shown in Figure 21 in terms of the action of the pivot. More specifically, in Fig. 21, the coupling link 33 is generally in compression, and the force-modifying mechanism pushes on the movable part 20 via the coupling link 33 so as to apply the actuating force to the movable part 20. Conversely, in Fig 22, the coupling link 33 is generally in tension, and the force-modifying mechanism pulls on the movable part 20 via the coupling link 33 so as to apply the actuating force to the movable part 20. Tension is particularly preferable, for example, when the coupling link 33 comprises a flexure. Figure 23 is a schematic plan view of another actuator assembly 2. This configuration allows for rotational movement of the movable part 20 relative to the support structure 10 about the primary axis. Optionally, the flexure arrangements 72 are provided over two planes, and may be etched from two sheets of metal. Figure 24 is a schematic plan view of another actuator assembly 2. This configuration allows for rotational movement of the movable part 20 relative to the support structure 10 about the primary axis. Optionally, the flexure arrangements 72 are provided using a single etching, i.e. from a single sheet of metal. OptionaIly, the length of the coupling flexure 33 may be increased to be greater than the length shown in Figure 24. Figure 25 is a schematic plan view of another actuator assembly 2. This configuration comprises an forcemodifying flexure 32 and a coupling flexure 33 held generally in tension combined with a body portion 31 held generally in compression. "Folded wire" examples Figure 26 is a schematic diagram illustrating an SMA wire 34 which is folded around a feature 57 (e.g. a pulley) such that the SMA wire 34 includes first and second lengths orientated at a non-zero angle relative to each other and these lengths are coupled to each other such that contraction of the first length causes a displacement of the second length that provides a contribution to movement of the movable part that is in addition to the contribution provided by contraction of the second length. Further details are provided in WO2021 / 111131 Al which is incorporated herein by this reference. Any of the configurations described herein may be combined with such folded SMA wires 34 in order to provide additional stroke, as shown in Figure 26. Other variations It will be appreciated that there may be many other variations of the above-described examples. For example, the flexure arrangements 72 shown in Figure 16 may be provided with an integrated crimp 35 as shown in Figure 18. The actuator assembly 2 shown in Figure 16 may be provided with a bearing 55 on one or both sides of the flexure arrangements 72. When only one side is provided with the bearing 55, the flexures 34, 33 and / or the SMA wire 34 may be preloaded or may extend slightly along the direction of the primary axis so as to provide a biasing force for the bearing 55, for example as shown in Figure 18. The arrangement shown in Figure 16 may be modified such that the body portion 31 acts as a class 3 lever instead of as a class 1 lever. Instead of being used in a camera assembly, the actuator assembly may be used in other devices. For example, the actuator assembly 2 may correspond to (part of) an illumination source which may be for use in a 3D sensing system such as described in WO2020 / 030916 or in an augmented reality (AR) display system, or the actuator assembly 2 may be used to move a display. SMA element The above-described SMA actuator assemblies comprise at least one SMA element. The term 'shape memory alloy (SMA)element' may refer to any element comprising SMA. The SMA element may be described as an SMA wire. The SMA element may have any shape that is suitable for the purposes described herein. The SMA element may be elongate and may have a round cross section or any other shape cross section. The cross section may vary along the length of the SMA element. The SMA element might have a relatively complex shape such as a helical spring. It is also possible that the length of the SMA element (however defined) may be similar to one or more of its other dimensions. The SMA element may be sheet-like, and such a sheet may be planar or non-planar. The SMA element may be pliant or, in other words, flexible. In some examples, when connected in a straight line between two components, the SMA element can apply only a tensile force which urges the two components together. In other examples, the SMA element may be bent around a component and can apply a force to the component as the SMA element tends to straighten under tension. The SMA element may be beam-like or rigid and may be able to apply different (e.g. non-tensile) forces to elements. The SMA element may or may not include material(s) and / or component(s) that are not SMA. For example, the SMA element may comprise a core of SMA and a coating of non-SMA material. Unless the context requires otherwise, the term 'SMA element' may refer to any configuration of SMA material acting as a single actuating element which, for example, can be individually controlled to produce a force on an element. For example, the SMA element may comprise two or more portions of SMA material that are arranged mechanically in parallel and / or in series. In some arrangements, the SMA element may be part of a larger SMA element. Such a larger SMA element might comprise two or more parts that are individually controllable, thereby forming two or more SMA elements. The SMA element may comprise an SMA wire, SMA foil, SMA film or any other configuration of SMA material. The SMA element may be manufactured using any suitable method, for example by a method involving drawing, rolling, deposition and / or other forming process(es). The SMA element may exhibit any shape memory effect, e.g. a thermal shape memory effect or a magnetic shape memory effect, and may be controlled in any suitable way, e.g. by Joule heating, another heating technique or by applying a magnetic field.

Claims

1. An actuator assembly comprising:first and second parts that are movable relative to each other in a movement plane that is perpendicular to a primary axis; andat least one pair of actuating units, each actuating unit connected between the first and second parts and arranged to apply a respective actuating force between the first and second parts,wherein one of each pair of actuating units comprises:a body portion;an SMA element connected between the body portion and the first part, and arranged, on actuation, to apply an input force to the body portion;a force-modifying element connected between the body portion and the first part and configured to modify the input force so as to give rise to the actuating force; anda coupling link connected between the body portion and the second part, wherein the coupling link is configured to transmit the actuating force from the body portion to the second part, and wherein the coupling link is compliant in a direction perpendicular to the direction of the actuating force; andwherein the other of each pair of actuating units comprises:a body portion;an SMA element connected between the body portion and the second part, and arranged, on actuation, to apply an input force to the body portion;a force-modifying element connected between the body portion and the second part and configured to modify the input force so as to give rise to the actuating force; anda coupling link connected between the body portion and the first part, wherein the coupling link is configured to transmit the actuating force from the body portion to the first part, and wherein the coupling link is compliant in a direction perpendicular to the direction of the actuating force.

2. An actuator assembly according to claim 1, wherein none of the actuating units overlap when viewed along the primary axis.

3. An actuator assembly according to claim 1 or 2, comprising a total of four actuating units, wherein the four actuating units consist of two pairs of actuating units.

4. An actuator assembly according to any preceding claim, wherein the four actuating units are in an arrangement capable of applying actuating forces between first and second parts so as to move thefirst and second parts relative to each other to any positions within a range of movement within the movement plane without applying any net torque between the first and second parts.

5. An actuator assembly according to any one of the preceding claims, comprising an image sensor 5 and / or a lens assembly, wherein the image sensor or the lens assembly is fixed relative to the second part and / or the lens assembly or the image sensor is fixed relative to the first part.

6. A camera assembly comprising an actuator assembly according to any one of the preceding claims, wherein the second part comprises an image sensor as an element of the imaging system, the 10 image sensor having an imaging axis perpendicular to the movement plane, wherein the actuator assembly provides optical image stabilisation (OIS) for the camera assembly.30

Images