GB2531075A

GB2531075A - Smart trolley wheel

Info

Publication number: GB2531075A
Application number: GB1418001.2A
Authority: GB
Inventors: Mark Denison Robert
Original assignee: Cambridge Consultants Ltd
Current assignee: Cambridge Consultants Ltd
Priority date: 2014-10-10
Filing date: 2014-10-10
Publication date: 2016-04-13
Also published as: GB201418001D0; WO2016055815A1

Abstract

A position tracking system is disclosed which derives information relating to a location of a cart or trolley, such as a shopping trolley in a supermarket. The cart comprises at least one wheel that is configured to acquire and output information representing a rotational distance travelled by the wheel and to determine and output a bearing of the wheel during motion, for example using an encoder, gyroscope, magnetometer, accelerometer or similar sensors. The system obtains information relating to an estimated change in a position of the cart and identifies, based on the wheel information, at least one location at which the cart was located. The wheel may also include means for wireless communication and wireless signal triangulation positioning may be used. Position estimates may include generating candidate state particle sets. Specific locations may be identified such as checkouts, aisles, near items of interest and restricted areas, for example to prevent the trolley being stolen.

Description

Smart Trolley Wheel The present invention relates to location tracking and in particular to tracking customer location in a supermarket environment, other retail environments, or the like. The invention has particular although not exclusive relevance to an apparatus for location finding and tracking using an intelligent wheel for a shopping trolley where conventional location finding apparatus, such as global positioning system (GPS) technology or the like cannot be used effectively (e.g. indoors).

Supermarkets and other retail businesses employ various methods for monitoring customers' shopping habits in order to be able to offer them personalised offers and in turn to increase sales. The most common way of monitoring shopping habits is by inviting customers to join loyalty schemes and to use a unique identifier (such as a customer number! loyalty card) at checkout.

Studies have shown that impulse purchases account up to 40% of all supermarket sales (even very conservative estimates put the figure at 20%). In response to these studies, retailers work hard to suggest items and make offers clear to customers at all stages of the shopping process. It is therefore important for retailers to know (apart from the goods purchased) what products their customers did not buy but which they may have considered buying and!or might be particularly interested in.

Using this insight, retailers are able to appeal to their customers' needs more efficiently and improve various aspects of their shopping experience. For example, retailers can further improve personalisation of offers for their customers, improve inventory handling! product availability, and generally augment their sales strategies based information about their customers. Whilst loyalty schemes provide a good understanding of customers' purchases, including impulse purchases that have been made by them, it is difficult for retailers to identify items that the customers did not buy but which they might be interested in buying in the future.

Other than costly and often ineffective customer surveys, a list of potential goods that a customer considered buying can be implied by tracking the customer's location within a store and by comparing the list of items purchased by that customer with the areas of the store (and any goods displayed in such areas) where the customer has spent time. Goods displayed at areas where a particular customer has spent some time but that do not appear on the customer's list of purchased items may be indicative of a missed sales opportunity. Retailers are thus very keen on identifying such items and turning any missed opportunities into a future sale (either through general and/or personalised offers).

There are systems that are designed to determine a customer's position within a particular store using cameras, which may be further enhanced in order to be able to distinguish between individual customers. However, these camera systems mainly focus on tracking heat maps of general customer activity within isolated areas of the store rather than individual customer tracking. Further, the more customers there are in a store, the more difficult it is for camera-based systems to distinguish one customer from another.

It is now commonplace for mobile telephones and other portable computational devices to include location finding and related features, such as Global Positioning System (GPS) for tracking a user's position, and electronic compasses in the form of magnetometers for direction finding. There are various systems that rely on navigation capabilities of customers' mobile telephones to track customer location and movement within a store, although GPS tracking is not possible in most retail environments (i.e. enclosedlindoor retail environments) and other features of the phone have to be used to enable such functionality. Such systems typically require customers to install and use a dedicated application on their mobile telephones.

However, it is difficult to ensure compatibility with every type of mobile telephone that the customers may use. Compatibility issues notwithstanding, such dedicated applications generally result in a relatively low level of engagement from the customer's part (as for each retailer there may be a separate, retailer-specific application that needs to be installed and running on the customer's mobile telephone). Wbilst such mobile telephone based tracking may be used to derive a list of potential items that the customer is interested in, user consent may often be required to the use of any tracking features provided by such applications. Even when the customer consents to tracking his/her location, in the absence of satellite (such as GPS) based input, the localisation tends to be unreliable (or even impossible) indoors to an accuracy required for many typical use cases.

Some of the drawbacks associated with such tracking solutions may be mitigated by using non-satellite based wireless capabilities (e.g. Bluetooth I Wi-Fi) of the customer's mobile telephone together with wireless transceivers installed at known locations within an area to be tracked (e.g. within a store). Such wireless transceivers may be referred to as ranging beacons' or simply beacons'. In this case, an application running on the customer's mobile telephone may determine the mobile telephone's (and hence the customer's) current position by triangulation, i.e. based on the distance between the mobile telephone and one or more of such beacons.

However whilst mobile phones can be used to triangulate position using signal strength measurements, this alone does not currently provide sufficient accuracy to determine detailed customer information (e.g. which products they spent significant time in front of). Indeed, the accuracy (-3m) achieved by current systems would likely struggle to confirm even which aisle a customer was in.

Further, such approaches also do not overcome other drawbacks generally associated with mobile telephone based solutions, such as a low level of user engagement arising, for example, from the potential need to obtain user consent to tracking, from a reluctance by a user to install an application on their mobile phone for every store they use, and/or simply from a forgetfulness to switch on and use the application. Generally, customers are becoming more and more aware of privacy and, as a result, are increasingly likely to object to being tracked using their own hardware. Whilst asking permission to use a customer's own device for location tracking is not currently a legal requirement in many countries, tracking in the absence of consent can frustrate and annoy customers thereby decreasing customer experience quality and damaging retailer reputation. It is, of course, a possibility that permission will become a requirement in the future even where it is not already a requirement.

Furthermore, in order to maintain accurate position tracking using conventional wireless technologies (such as GPS/Bluetooth/Wi-Fi), a relatively significant amount of energy is needed. This is a particular issue for portable devices such as mobile telephones, in which it is important to keep energy usage as low as possible in order to preserve battery life. Thus, if a customer starts an app then it will likely consume battery resources on their device whilst tracking their location which would be a significant inconvenience to the customer thereby making the customer less likely to choose to turn the application on.

Moreover, such approaches do not address the difficulties arising from the need for compatibility with the many types of mobile telephones used by the customers.

Accordingly, preferred embodiments of the present invention aim to provide methods and apparatus which provide reliable customer tracking that overcomes or at least alleviates one or more of the above issues.

In one aspect, the invention discloses a cart for carrying items, the cart comprising at least one wheel comprising: means for acquiring information representing a rotational distance travelled by the wheel; and means for outputting information acquired by said acquiring means for estimating a change in a position of the cart based on said information acquired by said acquiring means.

The cart may further comprise means for determining a bearing of the wheel during motion, in which case the information outputting means may be operable to output information acquired by said acquiring means for estimating a change in a position of the cart based on said information acquired by said acquiring means and said bearing determined by said determining means.

The cart may further comprise means for estimating a change in a position of the cart based on said output information acquired by said acquiring means and a bearing determined by said determining means.

The cart may further comprise means for communicating wirelessly (e.g. using a Bluetooth and/or a Wi-Fi based communication technology) with at least one of: another cart; a localisation beacon; and server apparatus.

In one aspect, the invention discloses a wheel for the above described cart, wherein said wheel comprises said means for acquiring information representing a rotational distance travelled by the wheel.

The wheel may further comprise means for determining a bearing of the wheel during motion, in which case the information outputting means may be operable to output information acquired by said acquiring means for estimating a change in a position of the cart based on said information acquired by said acquiring means and said bearing determined by said determining means.

The wheel may further comprise means for estimating a change in a position of the cart based on said output information acquired by said acquiring means and a bearing determined by said determining means.

The information acquiring means may comprise at least one sensor. In this case, the at least one sensor may comprise at least one of: a rotary encoder, a gyroscope, a magnetometer, and an accelerometer.

The wheel may further comprise means for communicating wirelessly with at least one of: another wheel of a cart; a localisation beacon; and server apparatus. The communicating means may be operable to communicate using a Bluetooth and/or a Wi-Fi based communication technology.

The wheel may further comprise means for generating power for powering at least said information outputting means. The power generating means may be operable to generate power during rotation of said wheel.

The wheel may further comprise means for refining an estimated position of the cart using triangulation based on a strength of signals received by said communicating means from a plurality of localisation beacons.

The wheel may further comprise means for refining an estimated position of the cart by generating a particle set comprising a plurality of particles each particle representing a respective estimate of a candidate state of the cart, wherein the estimated candidate state comprises an estimated position of the cart and at least one estimated error associated with movement of said cart; and adapting the particle set to reflect an estimated change in position represented by an estimated distance travelled by the cart, and an estimated direction of movement of the cart.

In one aspect, the invention discloses a position tracking system for deriving information relating to a location of the above described cart, the apparatus comprising: means for communicating with the corresponding communicating means of the cart, to obtain information relating to an estimated change in a position of the cart; and means for identifying, based on said information obtained from the communicating means of the cart, at least one location at which the cart was located.

The position tracking system may comprise means for associating a user with the cart and for deriving information specific to said user based on said at least one identified location.

The at least one location may comprise at least one of: a location within a predetermined area (such as a checkout area and/or an aisle); a location adjacent to a known location of a specific item of interest (such as an item that has / has not been purchased by a user associated with the cart); a location indicative of a positioning error; a location adjacent to an area in which position measurements are not available (e.g. spillage); and a location within an area falling outside a predetermined perimeter.

The position tracking system may comprise: means for storing information identifying a plurality of locations of interest; and means for determining whether the cart is in and/or proximate one or more of the plurality of locations of interest.

The plurality of locations of interest may include the location of at least one of: a wireless beacon; a specified product; an aisle; an entry point to an area (e.g. a store entrance, aisle entrance, cart park and/or the like); an area perimeter (e.g. a security perimeter); a shelf; a wall; a checkout area; another cart; and/or a staff member.

The position tracking system may further comprise means for tracking a route taken by the cart.

In one aspect, the invention discloses a method of deriving positional information for a cart for carrying items, the method comprising: acquiring, at a wheel of the cart, information representing a rotational distance travelled by the wheel; and outputting information acquired by said acquiring step for estimating a change in a position of the cart based on said information acquired by said acquiring step.

In one aspect, the invention discloses a method performed by a position tracking system for deriving information relating to a location of the above described cart, the method comprising: communicating with the cart, to obtain information relating to an estimated change in a position of the cart; and identifying, based on said information obtained from the cart, at least one location at which the cart was located.

Aspects of the invention extend to computer program products such as computer readable storage media having instructions stored thereon which are operable to program a programmable processor to carry out a method as described in the aspects and possibilities set out above or recited in the claims and/or to program a suitably adapted computer to provide the apparatus recited in any of the claims.

Each feature disclosed in this specification (which term includes the claims) and/or shown in the drawings may be incorporated in the invention independently (or in combination with) any other disclosed and/or illustrated features. In particular but without limitation the features of any of the claims dependent from a particular independent claim may be introduced into that independent claim in any combination or individually.

Embodiments of the invention will now be described by way of example only with reference to the attached figures in which: Figure 1 schematically illustrates an environment in which embodiments of the present invention may be applied; Figure 2 is a block diagram of a smart trolley wheel that may be used in the environment shown in Figure 1; Figure 3 schematically illustrates apparatus for use with the smart trolley wheel of Figure 2 to locate and track a customer in an environment such as that illustrated in Figure 1; Figure 4 shows an example flow chart illustrating a method performed by a location tracking apparatus when deriving a personalised offer for a customer based on position data; and Figures 5 to 7 illustrate schematically further scenarios in which the smart trolley wheel of Figure 2 may be used.

Overview Figure 1 schematically illustrates an environment, such as a supermarket, in which embodiments of the present invention may be applied. The supermarket includes an area 1 in which a customer 5 is to be tracked, and in which various products 7, 8 are displayed. The customer 5 has an associated shopping trolley 10 (shopping cart) thus the location of the customers can be derived with sufficient accuracy by locating his/her shopping trolley 10. It will be appreciated that in the scenario shown in Figure 1 only one customer 5 and one shopping trolley 10 are shown for illustration purposes, in real life scenarios a plurality of trolleys and a plurality of customers may be present within the store represented by area 1.

In this example, a number of wireless (e.g. Bluetooth) beacons 35 are installed at known positions throughout the area 1 of the supermarket in order to facilitate locating and tracking the customer 5. However, instead of relying on a conventional camera or mobile telephone based tracking solution for locating and/or tracking the customer 5, in this case the position of the associated shopping trolley 10 is being tracked.

This is realised by equipping the shopping trolley 10 with an intelligent wheel 20 which is configured to carry out position measurements by communicating with one or more of the beacons 35 over a radio frequency (RF) link, such as a Bluetooth and/or a Wi-Fi link. Specifically, the trolley wheel 20 is configured to determine its current location (and thereby approximate the current position of the customer 5) by performing an appropriate triangulation procedure, e.g. based on (known) positions of the beacons 35-1 to 35-4 that are within communication range of the wheel 20 and the respective distances (shown in dotted lines in Figure 1) between the trolley wheel and the beacons 35-1 to 35-4. It will be appreciated that each beacon 35 may transmit its own position and/or their position may be determined from information stored by (and/or made available to) the trolley wheel 20. Distance between the trolley wheel 20 and a particular beacon 35 can determined, for example, based on the strength of a signal received from that beacon 35 and/or based on the time when a signal transmitted by the beacon 35 reaches the trolley wheel 20 (assuming the time of transmitting the signal is known).

The trolley wheel 20 is also configured to transmit (via beacons 35) information about the trolley's 10 position to a remote server (not shown in Figure 1), which server is configured to compare the location of the trolley 10 to a database of products (or product types) that are found at that location and/or in its vicinity. The information transmitted by the trolley wheel 20 may include, for example, the position determined by the trolley wheel 20 (e.g. the distance(s) from one or more beacon and/or the measured position) at a given time, and this information may be transmitted substantially continuously and/or when a predetermined condition is met (e.g. the trolley 10 is at checkout).

It will be appreciated that triangulation allows the trolley wheel's 20 position to be determined when its distance from three different beacons 35 (and in any case from at least one beacon 35) is known. Position accuracy may be improved by receiving signals and obtaining distances from more than three beacons (e.g. beacons 35-1 to 35-4 shown in Figure 1).

The trolley wheel 20 is also equipped with one or more sensors, such as a rotation sensor (rotation encoder) for improving the location accuracy that can be achieved by triangulation alone. For example, the wheel 20 may use one or more rotation sensors for measuring: i) rotation of the wheel 20 for determining a distance travelled (based on the diameter and/or circumference of the wheel 20 and the number of rotations completed); and ü) rotation/orientation of the wheel 20 about an axis of a leg of the trolley 10 that it is attached to for determining a yaw of the wheel 20 as the customer pushes the trolley 10.

In this case, using one or more of such sensors, the trolley wheel 20 is configured to track the distance (and possibly direction) travelled by the trolley 10 and to track a yaw of the wheel 20 in relation to the trolley 10. Such tracking is employed by the trolley wheel 20 for the purposes of a so-called dead reckoning'. Dead reckoning is a positioning technique using which a current position may be obtained based on a previously known' position or fix' in combination with estimates of distance and direction travelled from that position. In this example, a known position (or fix') may be obtained using any suitable method, such as triangulation and/or the like.

It will be appreciated that, over time, the location calculated from the data obtained from the sensors may introduce a discrepancy with the actual location of the trolley within the area 1. However, such discrepancy may be reduced significantly by using at least two positioning methods (e.g. both triangulation and dead reckoning) for determining the location of the trolley 10. If location information is gathered and processed in real time and using two positioning methods, accuracy of the location information is expected to be within lm, which is likely to be sufficient for the purposes of tracking customers within a supermarket and/or the like.

The trolley wheel 20 may also be configured to obtain its location and/or to further improve its location accuracy by communicating, via the beacons 35, with the remote server (and/or the like), to which the trolley wheel 20 may send information relating to the position obtained by triangulation/dead reckoning. In this case, the remote server may be configured to compare the received information (trolley position) with a database comprising a layout of the area 1 and/or information on the location of products 7, 8. The server may also be configured to discard/eliminate any positions obtained by triangulation/dead reckoning that are considered to be unrealistic, for example, positions falling outside the area 1. If applicable, the trolley wheel 20 may also be configured to obtain its location based on information (e.g. a layout of the area 1 and/or location of the beacons 35) received from the remote server and/or from other data sources. It will be appreciated that such remote server based location tracking techniques may be used, when appropriate, in combination with any triangulation and/or dead reckoning performed by the trolley wheel 20.

Advantageously, the smart wheel 20 in this example also comprises a generator (and an associated baftery) so that power required by the wheel's electronic components (sensors, transceiver, etc.) can be provided without requiring the smart wheel 20 to be connected to an external power source. For example, a direct current (DC) generator may be provided inside the smart wheel 20 for generating the required power.

In the scenario shown in Figure 1, the location of the trolley 10 is initially determined to be in the vicinity of product 7. Accordingly, it can be assumed with some certainty that the customer 5 considers purchasing this product 7 (e.g. if the trolley 10 is determined to remain in the vicinity of product 7 for at least a predetermined amount of time and/or for a number of consecutive triangulation rounds).

It will be appreciated that based on a series of consecutive position measurements, the path of the customer 5 (i.e. the path of the customer's shopping trolley 10 to which the smart wheel 20 is attached) can be determined within the area 1 with sufficient accuracy to allow identification of a set of products and/or product types that the customer 5 is interested in. For example, as the customer 5 moves down an aisle, the location of the trolley 10 may be determined (e.g. in a subsequent triangulation / dead reckoning round) to be in the vicinity of product 8, in which case the product 8 may also be included in the set of products (which already include product 7) that the customer 5 is interested in.

The trolley 10 used by the customer S can be uniquely identified by comparing which customer is being served when the trolley wheel's 20 location is known to be adjacent to the checkout area. Thus, when the customer 5 completes his/her checkout, it is possible to determine whether the customer 5 has purchased either one of the identified products 7,8 near which the customer's 5 trolley 10 was located (or stopped for a predetermined amount of time).

Therefore, it is possible for the retailer to gather information on the customer's 5 shopping habits, including identifying any products that the customer 7 only considered but did not buy, and to improve personalisation of offers for the customer based on this information.

Beneficially, there is no need to provide and install any software applications on the customer's equipment (such as their mobile telephone) since the smart wheel 20 forms part of the retailer's own shopping trolley 10. This way, since the location of the shopping trolley 10 is being tracked, there is no need to obtain user permission for doing so (although the customers may be advised that such tracking is taking place).

Further, since the customer's 5 trolley 10 can be uniquely identified latest during the checkout process (and linked to the customer's 5 profile in the retailer's database), any items purchased by that customer 5 can be compared to a list of products (derived based on position data) that the customer 5 potentially considered buying.

Even if the customer 5 cannot be identified (e.g. because the customer 5 does not have a loyalty card), it is still possible to gather valuable information on his/her shopping habits and journey taken through the store.

Since the trolley wheel 20 is equipped with its own power source (generator/battery), once the beacons 35 are installed and in operation, the system is able to track the customer's 5 shopping habits reliably and autonomously, requiring a relatively low level of maintenance (if any).

Smart trolley wheel Figure 2 is a block diagram illustrating the main components of the smart trolley wheel 20 shown in Figure 1.

As shown, the wheel 20 includes a processor 11, and an associated memory 12.

Software may be pre-installed in the memory 12 and/or may be downloaded, for example, via a telecommunications network, via a Wi-Fi/Bluetooth link (e.g. via beacons 35), and/or from a removable data storage device (RMD).

The wheel includes communications circuitry 13, for example a Bluetooth/Wi-Fi transceiver and/or the like, for communications with other devices (such as the beacons 35) via an antenna 14.

The processor 11 is configured to control the overall operation of the smart wheel 20 by, in this example, program instructions or software instructions stored within the memory 12. As shown, these software instructions include, among other things, an operating system 15, a communications control module 16, a triangulation module 17; a dead reckoning module 18, and a position filtering module 19.

The communications control module 16 controls communication with the beacons 35 including, for example, communication relating to triangulation measurements. The communications control module 16 also controls communication with remote customer tracking apparatus 30 (e.g. via the beacons 35).

The triangulation module 17 is responsible for deriving distances between the trolley wheel 20 and one or more beacons 35 within communication range of the trolley wheel 20. If the positions of the beacons 35 are known (e.g. stored in local memory 12 and/or transmitted by the beacons 35), it is possible to derive a current position of the trolley wheel 20 following an appropriate triangulation procedure based on the beacon's 35 locations and the corresponding distances. The triangulation module 17 may be configured to perform such a triangulation procedure itself (e.g. if the beacon locations are known), although the triangulation module 17 may also be configured to send the derived distances to the customer tracking apparatus 30. In the latter case, the customer tracking apparatus 30 may perform a triangulation procedure in order to derive the trolley wheel's 20 location.

The dead reckoning module 18 is configured to estimate (based on data obtained from one or more sensors 25 to 28) the trolley wheel's 20 current position by tracking the distance (and possibly direction) travelled by the trolley 10 from a previously known position.

The position filtering module 19 is responsible for deriving the most likely position of the trolley wheel 20 (and/or for eliminating position values that are likely to have an error) using statistical methods and/or data relating to a layout of the area 1. For example, the position filtering module 19 may obtain respective position values from both the triangulation module 17 and the dead reckoning module 18, and determine which one of the obtained position values is more likely to be correct (or have better accuracy).

The trolley wheel 20 includes a generator 21 for generating power from the movement (rotation) of the wheel 20. The generator is coupled to a battery 22 for storing energy and for powering the wheel's various electrical components whilst the generator 21 does not generate power (e.g. when the trolley wheel 20 is stationary).

The trolley wheel 20 also includes one or more sensors, such as one or more rotary encoders 25, one or more gyroscopes 26, one or more magnetometers 27, and one or more accelerometers 28. It will be appreciated that any of these sensors 25 to 28 may be optional (hence they are shown in dashed lines in Figure 2).

Customer tracking apparatus Figure 3 schematically illustrates apparatus 30 for use with the smart trolley wheel 20 of Figure 2 to locate and track a customer 5 in the area 1 illustrated in Figure 1.

As shown, the customer tracking apparatus 30 includes a server 31 coupled to a map data source 33. The map data source 33 includes information relating to the layout of the area 1 in which the customer 5 is to be located, and/or information identifying the respective positions of beacons 35 within that area 1.

The server 31 is also connected to a number of beacons 35, using any suitable wired and/or wireless technology. Each beacon 35 is equipped with wireless communication capability (e.g. Bluetooth/Wi-Fi and/or the like) for communicating with the trolley wheel 20.

The server 31 includes a processor, and an associated memory 40. Software may be pre-installed in the memory 40 and/or may be downloaded via a communications network or from a removable data storage device (RMD), for example. The processor is configured to control overall operation of the server 31 by, in this example, program instructions or software instructions stored within the memory 40. As shown, these software instructions include, among other things, an operating system 41, a communications control module 42, a (trolley) positioning module 43; a position filtering module 44, and an inventory module 45.

The communications control module 42 controls communication with the trolley wheel 20 (via the beacons 35) including, for example, communication relating to triangulation measurements. The communications control module 42 also controls communication with the map data source 33.

The positioning module 43 is responsible for deriving the trolley wheel's 20 position, for example, by performing an appropriate triangulation procedure based on the beacon's 35 locations and the corresponding distances between the beacons 35 and the trolley wheel 20. In some cases, the trolley wheel 20 may be configured to perform such a triangulation procedure itself (e.g. if the beacon locations are known), in which case the positioning module 43 obtains the result of the triangulation procedure from the trolley wheel 20 (via beacons 35). Alternatively, the positioning module 43 may be configured to obtain the derived distances (between the trolley wheel 20 and one or more beacon 35) from the trolley wheel 20, in which case the positioning module 43 may be configured to perform a triangulation procedure in order to derive the trolley wheel's 20 position.

The position filtering module 44 is responsible for deriving the most likely position of the trolley wheel 20 (and/or for eliminating position values that are likely to have an error) using statistical methods. For example, the position filtering module 45 may obtain a set of candidate position values from the trolley wheel 20 (derived by e.g. an appropriate triangulation/dead reckoning procedure) and/or from the positioning module 43, and determine which one of the obtained position values is more likely to be correct (or have better accuracy).

The inventory module 45 includes a database of items (such as products 7 and 8) and their associated positions within the area 1. The inventory module 45 is responsible for deriving a list of items that the customer 5 is likely to be interested in, for example, based on the route the customer 5 has taken within the area 1. The route of the customer 5 is obtained based on the position data provided by the other modules, such as the positioning module 43 and/or the position filtering module 44.

The inventory module 45 is configured to associate the trolley 10 with the customer 5 (e.g. at checkout) and compare the list of items purchased by that customer 5 with the list of items that the customer 5 is determined to be interested in. If there are any items that the customer 5 is likely to be interested in but that the customer 5 did not purchase, the inventory module 45 is configured to mark such un-purchased items as candidates for a personalised offer for the customer 5.

In the above description, the trolley wheel 20 and the customer tracking apparatus 30 are described for ease of understanding as having a number of discrete modules (such as the communication control modules, the triangulation modules, and the dead reckoning module). Whilst these modules may be provided in this way for certain applications, for example where an existing system has been modified to implement the invention, in other applications, for example in systems designed with the inventive features in mind from the outset, these modules may be built into the overall operating system or code and so these modules may not be discernible as discrete entities. These modules may also be implemented in software, hardware, firmware or a mix of these.

Operation Figure 4 shows an example flow chart illustrating, by way of example only, a method performed by the location tracking apparatus 30 when deriving a personalised offer for a customerS based on position data obtained from the smart trolley wheel 20.

The process starts at step S100. In step SlOl, the location tracking apparatus 30 (using its positioning module 43) obtains the trolley's 10 current position from the trolley wheel 20.

Next, in step S102, the location tracking apparatus 30 determines (using its position filtering module 45) whether or not the obtained position is useable. For example, the position filtering module 45 may check (e.g. by comparing the obtained position and a map of the area 1 provided by the map data source 33) whether or not the obtained position is within the area 1. If the position is outside the area 1, then it is determined to be not useable for this procedure. The position filtering module 45 may also check whether or not the obtained position shows discrepancy with position data obtained from other sources (or with previous position data), which may be indicative of a positioning error. Such a positioning error may be present e.g. due to interference and/or the like. If the discrepancy is found to be above a threshold, then the obtained position data is determined to be not useable for this procedure.

If the position data is determined to be not useable (step 5102: NO'), then the location tracking apparatus 30 proceeds to step S103 in which it discards the position data. The procedure then returns to step SlOl in order to obtain a subsequent position data for the trolley wheel 20. Although not shown in Figure 4, when a position data is discarded, the location tracking apparatus 30 may also notify (e.g. via the beacons 35) the trolley wheel 20 about this problem so that the trolley wheel 20 is able to re-calibrate its positioning capabilities (its triangulation module 17 and/or its dead reckoning module 18) accordingly.

If the position data is determined to be useable (step 5102: YES'), then the location tracking apparatus 30 proceeds to step S104 in which it updates (using its inventory module 44) a list of items that can be correlated to the obtained position data for this trolley wheel 20. Specifically, the inventory module 44 compares the location of each product 7, 8 (based on a map of the area 1 provided by the map data source 33) to the location obtained for the trolley 20 (and hence the customer 5). Any product within a predetermined vicinity of the obtained position (in this example, product 7) may be added to a list of candidate items for inclusion in a personalised offer for the customer 5 associated with that trolley wheel 20 (although it will be appreciated that such association between the customer 5 and the trolley wheel 20 may be made at a later phase).

Next, in step S105, the location tracking apparatus 30 checks whether the customer 5 is at checkout. This may be determined, for example, by the positioning module 43 determining that the trolley wheel 20 is located at a checkout section of the area 1.

Such a checkout section may be identified, for example, based on the map of the area 1 provided by the map data source 33. The location tracking apparatus 30 may also be configured to determine (or confirm) that the customer 5 is at checkout by communicating with an electronic point of sale (EPOS) system.

If the check at step S105 indicates that the customer 5 (i.e. the trolley wheel 20) is not yet at checkout (step S105: NO), then the procedure returns to step SlOl in which a subsequent position data is obtained.

However, if the check at step S105 indicates that the customer 5 (i.e. the trolley wheel 20) is at checkout (step 5105: YES), then the location tracking apparatus 30 proceeds to step 5106, in which it checks (using its inventory module 44) whether there are any items that have been added (at step S104) to the item list associated with this customerS.

If the check at step S106 indicates that there are no un-purchased items (i.e. the customer 5 purchased each previously identified product), then the procedure ends (as shown in step S108).

However, if the check at step S106 indicates that there are un-purchased items (step S106: YES'), then the location tracking apparatus 30 proceeds to step S107, in which it generates (using its inventory module 44) a personalised offer for the customer 5 associated with the trolley wheel 20. For example, if it is determined that the customer did not purchase product 7 (which was added to the item list at step S104), then the inventory module 44 may include product 7 in a subsequent offer for this customers.

The procedure ends at step 5108, at least until the trolley 10 is taken back into the area 1 (e.g. by another customer).

Thus, advantageously, the above described location tracking apparatus is able to gather information on customers' shopping habits and identify any products that a particular customer is potentially interested in but that he/she but did not buy. Based on this information, it is possible to improve personalisation of offers for customers and thereby contribute to an increase of revenue and customer loyalty.

Modifications and alternatives Detailed embodiments have been described above. As those skilled in the art will appreciate, a number of modifications and alternatives can be made to the above embodiments whilst still benefiting from the inventions embodied therein.

In the example given above with reference to Figure 4, the system is described to collect customer location information for the purposes of identifying items that the customer may be interested in (but which he/she did not purchase). However, it will be appreciated that the system may be used for collecting other types of information and/or to implement other use cases as well. For example, the information collected may include information on one or more of the following: which aisles the customer did or did not visit (and optionally, why / why not visited a specific aisle); which end of the store did the customer start at; did the customer avoid aisles with lots of people in; what percentage of the store did the customer visit (i.e. before heading to the checkout); did the customer seemingly browse around the store and/or go to one or more specific locations quickly and then immediately to the tills; and/or the like. The information may be collected on a per customer basis and/or in an aggregated form (e.g. per time of day, per customer category, per promotion, per items purchased).

It will be appreciated that the location tracking apparatus may be configured to aggregate position data from the smart trolley wheels and provide real-time feedback to retailer staff in store and/or in a remote team. For example, the gathered information may be used in a number of areas, such as user journey tracking, store management, incident detection, gathering trolley lifecycle information, trolley theft reduction, etc. Store management / incident detection Figure 5 illustrates an exemplary scenario in which positioning techniques using smart trolley wheels 20 may be employed. Specifically, in this scenario, position data obtained (in substantially real-time) from a plurality of shopping trolleys 10 indicates that a particular area is avoided by the customers and/or that customers are less likely to spend time in that particular area (indicated by dotted lines in Figure 5).

Such location information may be used to provide a new level of detail for improving customer experience. For example, such information may be indicative of an incident such as the presence of a spillage and/or other type of contamination in the area 1.

Such position information (i.e. the lack thereof) may also be used to determine which products (or product types) need to be re-filled. Therefore, this information may be used in real time to optimise store management. Further examples include management of numbers of till assistants and other personnel in order to rapidly respond to changes in levels of activity in areas of the store from shelf refilling to management of offers.

Trolley lifecycle information With shopping trolleys costing over £100, any methods for lengthening their life would be valuable. It will be appreciated that the smart trolley wheel 20 may also be configured to monitor its state (e.g. any errors, overall distance covered, etc.) and/or the remaining lifetime of its components. In particular, the trolley wheel 20 may be configured to determine whether or not the trolley and/or the trolley wheel is damaged. Advantageously, such monitoring and/or determination may be made using information obtained from one or more of the sensors 25 to 28.

The trolley wheel 20 may be configured to provide such information to a remote server periodically, upon request, and/or when a predetermined condition is met. For example, a damaged trolley 10 may be identified (e.g. based on its movement/usage pattern) and reported in real time thereby further improving customer experience.

Trolley theft reduction Figure 6 illustrates another exemplary scenario, in which positioning techniques using smart trolley wheels may be employed. In this case, the system detects when a shopping trolley 10 is being taken outside a perimeter, such as a perimeter defined by a number of beacons arranged around an area 1' (e.g. a car park).

It is estimated that theft of supermarket trolleys costs $800million worldwide per annum. To reduce this cost supermarkets employ a variety of processes and techniques to prevent theft however, these often result in inconvenience to the customer and increased costs for the retailer.

However, in this case a perimeter is defined by a plurality of beacons deployed around the store. The trolley's position may be reported by the smart trolley wheel in substantially real time in order to indicate when the trolley is being removed from the premises. Specifically, the trolley wheel may be configured to report when the current position of the trolley 10 is determined to be outside of the perimeter defined by the beacons. For example, when the trolley is determined to be on the other side of the perimeter defined by beacons 35x and 35y (e.g. at the location labelled B' in Figure 6), an alarm may be triggered so that the trolley 10 can be retrieved without delay.

Particle filtering Many existing solutions for indoor tracking rely on some component of infrastructure, such as a ranging beacon, or knowledge of radio signal distribution from external infrastructure (e.g. cellular base stations) in the building, to alleviate accumulation of error in the positioning measurements. Whilst the present invention also relies on such ranging beacons, this increases the cost and time required to set up such a system. Accordingly, the number of beacons is preferably kept at a minimum.

Therefore, in order to improve the trolley wheel's dead reckoning capabilities, a so-called particle filtering technique may be used, which will be described with reference to Figure 7.

In this example, the trolley wheel is equipped with one or more inertial sensors (such as gyroscopes, multi-axis accelerometers, and/or the like) of the sort commonly integrated into modern smart' devices including cellular phones. Such inertial sensors (in combination with one or more rotary encoders) may be used to provide, at a relatively low cost, motion data required for estimating indoor location. Each time a position measurement is taken, a number of possible trolley positions are determined by the trolley wheel (e.g. using its triangulation module, dead reckoning module, etc.) based on information obtained from the sensors.

However, a significant limitation of such inertial sensing based dead reckoning is that the sensor measurements can be noisy and errors accumulate in the process of processing the measured quantities to obtain the desired positional measurements.

Therefore, every measurement typically includes an error, which gets exponentially larger and larger as consecutive measurements rely on the result of previous measurements (e.g. in the case of dead reckoning). Therefore, when such measurement errors are also taken into account, each measured position can be more accurately represented as a set of potential' or candidate' positions, hereafter referred to as a particle set, instead of a single position.

Figure 7 illustrates (a part of) an exemplary path that a customer may take through the area 1. In this example, the trolley's position is known at the location A', based on which the trolley's subsequent position is estimated at locations B' and C'. As can be seen, the particle set at each location where position measurement is made (such as locations B' and C') is represented by a number of white dots, each dot representing a respective position resulting from a particular position measurement (also taking into account any error), with the resulting estimated position derived from that particle set represented by the letter X. Whilst in Figure 7 position measurements are taken at three locations A to C, it will be appreciated that location measurements may be taken by the trolley wheel more often and/or substantially continuously, if appropriate.

At each location, the actual position of the trolley wheel is derived by averaging (taking the mean of) the various particles for that location after applying an appropriate filtering to the particle set.

Specifically, the set of particles comprises a plurality of particles, each particle representing a respective estimate of a candidate state of the trolley wheel, wherein the estimated candidate state comprises an estimated position of the trolley wheel.

Movement of the trolley wheel is detected and analysed to estimate distance travelled by the trolley and to determine an estimated direction of movement of the trolley and the particle set is updated to reflect an estimated change in position represented by the estimated distance travelled by the trolley wheel, and the estimated direction of movement of thereof.

Beneficially, the smart trolley wheel may be configured to refine its position estimates using a particle filtering technique to help reduce the build-up of systematic errors in the position estimates, as the trolley moves through the area. Particle filtering is a technique for estimating a state that changes over time using a sequence of potentially noisy measurements. It works by approximating the probability distribution for the state by a set of particles.

The particle filtering technique of this example advantageously uses a multi-dimensional state-space' with system states representing not only an estimated position of the trolley wheel but also an estimate of the errors in the motion measurements that resulted in that estimate position. In the illustrated example, the particle filtering technique uses a four-dimensional dimensional state-space' (rather than a simple two-dimensional state-space), for movement in two-dimensions (i.e. on substantially the same level), with each state' being represented by a position in an east-west direction (x'), a position in a north-south direction (y'), a bearing error and a wheel distance error.

A multi-dimensional probability density function associated with the estimated initial position (e.g. based on a known error distribution associated with the conventional positioning capability) is sampled to produce a set of candidate states for the estimated initial position where each candidate state is treated as a distinct particle.

Each particle is weighted, in this initial state, according to the probability associated with that particle from the probability density function taking account of any pre-existing knowledge about the initial position or the measurements from which the initial position was estimated (such as location A' in Figure 7) and/or uncertainty levels/error distributions associated with a conventional positioning measurement.

The sum of all weights attributed to all particles is equal to unity.

When the trolley moves to the second location B', the trolley wheel is able to detect and classify the movement from the motion data output from the inertial sensors. For example, the trolley wheel may be able to use this motion data to determine, in combination with a motion profile representing an expected average distance covered with each revolution of the trolley wheel, to allow the distance moved to be estimated. The trolley wheel is also able to determine, from the motion data output from the inertial sensors, an estimated bearing. Thus, the trolley wheel is able to predict a change in position for each particle based on the estimated distance covered by the wheel, bearing and associated error estimates based on the inertial sensing data.

Thus, at location B', the trolley wheel updates the respective weight of each particle based on any available measurement data (including measurement data from conventional positioning features if available). Advantageously, even where such measurement is not available, the trolley wheel is still able to make inferences about the likelihood that the movement of a particular particle and hence its current position is genuine based on stored mapping data for the area in which the trolley is being used. If a particle (such as the particles within the dotted lines in Figure 7) is found to have passed through an obstacle represented by such mapping data, such as a wall or item of fixed furniture, the particle's weight can be reduced to zero (or near zero) to indicate that that particle is very unlikely to represent the true position of the trolley wheel. Similarly, if a particle is found to take a lower probability route (e.g. very close to the edge of a dooiway rather than through its centre) then its weight can be reduced commensurately. The trolley wheel is also able to update a particle by adjusting the position and/or errors (state) associated with that particle to move it to a state that is more likely to be correct (this can be instead of, or in addition to, adjusting its weight).

The modified probability density represented by the re-weighted particle distribution may be re-sampled to remove low weight particles and to spawn new particles having a state close to that of the highest weight particles. For example, the new re-sampled' particle set resulting from the re-sampling process may have the same overall number of particles as were included in the previous set of particles. Each particle in the re-sampled paiticle set is then assigned a new, equal, weight whilst maintaining the unity sum of all particle weights (hence the weight of each particle is equal to the reciprocal of the number of particles in the paiticle set).

The updated particle set is used to determine a state estimate for the second location (e.g. location B' in Figure 7) by calculating, in this example, the mean over the set of particles. A measure of covariance can also be determined directly from the updated particle set in a straightforward way.

RF technology Whilst in the above description a Bluetooth based communication technology is used for illustrative purposes, it will be appreciated that the smart trolley wheel and the beacons may use any other suitable communication technology, such as Wi-Fi, Bluetooth Low Energy (BLE), and/or the like.

It will be appreciated that the number of beacons installed within the area in which customers need to be tracked may be limited. However, it will be appreciated that there is no need for the trolley wheel and the beacon to be able to communicate directly with each other. Further, it will be appreciated that communication between the trolley wheel and a beacon may be provided in a hop-by-hop fashion, i.e. via other trolley wheels. It will be appreciated that the trolley wheels and the beacons may be configured to communicate with each other in a mesh like network.

It will be appreciated that the beacons may be provided in a way (e.g. as part of a mesh infrastructure) so as to assure there are no dead spots in their coverage, especially during periods of low numbers of customers present in the store, which is a common problem when using a mesh architecture.

Triangulation based on distance from other trolley wheels It will be appreciated that the trolley wheel may be able to determine its position by triangulation even if all beacons are located outside the communication range of the trolley wheel. This may also be useful in scenarios when there are beacons within range of the trolley wheel but such beacons are not able to communicate with the trolley wheel directly (e.g. due to the beacons having reached a maximum number of simultaneous connections). It will also be appreciated that instead of using beacon locations, the current position of the trolley wheel may also be determined with respect to other trolleys (the location of which is known). In any case, it will be appreciated that the number of beacons required may not need to be proportional to the number of customers (trolleys) to be tracked. On the contrary, the larger the number of trolleys (having smart wheels), the larger positioning accuracy can be achieved even if the number of beacons is the same, because each smart trolley wheel may be used as a point of reference for the triangulation process.

In the above example, the beacons are described to implement a radio frequency (Bluetooth) technology for transmitting data based on which position calculations can be performed by the smart trolley wheel. However, it will be appreciated that such beacons may be implemented using any suitable data source capable of emitting at least one of non-satellite radio signals; visible electromagnetic radiation; and/or infra-red radiation.

Device placement It will be appreciated that using the wheel of the trolley (rather than a different part of the trolley) for position measurements has at least the following benefits: -ease of installation to existing trolleys (as a wheel replacement) and therefore reduced time for roll out, and/or cost of repairs/replacements; -the rotating wheel can be used as a power source; and -the ability to measure rotation (from wheel circumference) and yaw (angle to attachment to the trolley) of the wheel provides additional data to location tracking algorithms (e.g. for dead reckoning).

Smart Trolley Enhancements It will be appreciated that the smart trolley wheel based positioning system may be extended with additional technologies and/or features, which would add significant benefits to the customers.

For example, additional features may be realised by connecting the trolley wheel to a screen of some kind to display information. This can be provided in at least two ways: -through the use of a screen attached to the trolley itself; and -by coupling the trolley wheel to the customer's smart phone (e.g. via an appropriate Bluetooth connection and/or the like).

The screen (of the trolley and/or the customer's phone) may be used to indicate the location of items that the customer is trying to find and/or to offer an optimal route around the store to pick up the goods they need. This would bring one or more of the following potential benefits: -the customer's optimal route can be planned to make shopping more efficient and convenient; -the route can take into account all other shoppers present in the store (and/or their routes); -the route can take into account any busy areas in the store as the customer proceeds; -the screen may display details on the customer's current location (e.g. types of products, applicable product offers, direction of checkout/fire exit, and or the like); -it may be possible to gather feedback from the customer (e.g. via the screen) in order to assist item selection, for example by displaying premium products and/or other impulse purchase items in dependence on the customer's current location; -the screen may be used as a user interface (UI) for providing information about offers that are near to the customer's current location with a high level of accuracy -for example, not simply alcohol but red wine (e.g. red wine of a specific type and/or from a specific region); -the routing information may be used to allow the customer to be presented with suggestions which are not on their "usual" route around the store allowing the retailer to focus on items with which they have specific campaigns; and -based on an accurate location for the customer, the screen may be used for offering the ability for customers to report issues such as spillages and/or broken equipment in the store (possibly with a single button click).

In the above embodiments, a number of processing modules and processing steps were described. As those skilled in the art will appreciate, where the processing modules and/or steps described above are implemented in software the software may be provided in compiled or un-compiled form and may be supplied to the device as a signal over a computer network, or on a recording medium. Further, the functionality performed by part or all of this software may be performed using one or more dedicated hardware circuits.

In the above description various ways of tracking customer location in a supermarket environment are given. However, it will be appreciated that the invention is also applicable to any retail environment where trolleys/carts are used not only supermarkets. Further, it will be appreciated that embodiments of the invention may be applicable to tracking location (of a trolley and/or a person using a trolley) in non-retail environments as well, including any public or private space where trolleys are being used (e.g. airports, rail stations, hospitals, etc.).

Whilst the above embodiments have been described using a shopping trolley as an example, it will be appreciated that the embodiments are also applicable to any type of trolley/cart/barrow (and/or similar) that has at least one wheel.

Various other modifications will be apparent to those skilled in the art and will not be described in further detail here.

Claims

CLAIMS1. A cart for carrying items, the cart comprising: at least one wheel comprising: means for acquiring information representing a rotational distance travelled by the wheel; and means for outputting information acquired by said acquiring means for estimating a change in a position of the cart based on said information acquired by said acquiring means.
2. The cart according to claim 1, further comprising means for determining a bearing of the wheel during motion; and wherein said information outputting means is operable to output information acquired by said acquiring means for estimating a change in a position of the cart based on said information acquired by said acquiring means and said bearing determined by said determining means.
3. The cart according to claim 1 or 2, further comprising means for estimating a change in a position of the cart based on said output information acquired by said acquiring means and/or a bearing determined by said determining means.
4. The cart according to any of claims 1 to 3, further comprising means for communicating wirelessly with at least one of: another cart; a localisation beacon; and server apparatus.
5. A wheel for a cart according to claim 1, wherein said wheel comprises said means for acquiring information representing a rotational distance travelled by the wheel.
6. The wheel according to claim 5, further comprising means for determining a bearing of the wheel during motion; and wherein said information outputting means is operable to output information acquired by said acquiring means for estimating a change in a position of the cart based on said information acquired by said acquiring means and said bearing determined by said determining means.
7. The wheel according to claim 6, further comprising means for estimating a change in a position of the cart based on said output information acquired by said acquiring means and a bearing determined by said determining means.
8. The wheel according to claim 5, 6 or 7, wherein the information acquiring means comprises at least one sensor.
9. The wheel according to claim 8, wherein the at least one sensor comprises at least one of: a rotary encoder, a gyroscope, a magnetometer, and an accelerometer.
10. The wheel according to any of claims 5 to 9, further comprising means for communicating wirelessly with at least one of: another wheel of a cad; a localisation beacon; and server apparatus.
11. The wheel according to claim 10, further comprising means for refining an estimated position of the cart using triangulation based on a strength of signals received by said communicating means from a plurality of localisation beacons.
12. The wheel according to claim 10 or 11, wherein said communicating means is operable to communicate using a Bluetooth and/or a Wi-Fi based communication technology.
13. The wheel according to any of claims 5 to 12, further comprising means for generating power for powering at least said information outpufting means.
14. The wheel according to claim 13, wherein the power generating means is operable to generate power from rotation of said wheel.
15. The wheel according to any of claims 5 to 14, further comprising means for refining an estimated position of the cart by generating a particle set comprising a plurality of particles each particle representing a respective estimate of a candidate state of the cart, wherein the estimated candidate state comprises an estimated position of the cart and at least one estimated error associated with movement of said cart; and adapting the particle set to reflect an estimated change in position represented by an estimated distance travelled by the cart, and an estimated direction of movement of the cart.
16. A position tracking system for deriving information relating to a location of a cart according to claim 4, the apparatus comprising: means for communicating with the corresponding communicating means of the cart, to obtain information relating to an estimated change in a position of the cart; and means for identifying, based on said information obtained from the communicating means of the cart, at least one location at which the cart was located.
17. The position tracking system according to claim 16, comprising means for associating a user with the cart and for deriving information specific to said user based on said at least one identified location.
18. The position tracking system according to claim 16 or 17, wherein said at least one location comprises at least one of: -a location within a predetermined area (such as a checkout area and/or an aisle); -a location adjacent to a known location of a specific item of interest (such as an item that has / has not been purchased by a user associated with the cart); -a location indicative of a positioning error; -a location adjacent to an area in which position measurements are not available (e.g. spillage); and -a location within an area falling outside a predetermined perimeter.
19. The position tracking system according to any of claims 16 to 18, comprising: means for storing information identifying a plurality of locations of interest; and means for determining whether the cart is in and/or proximate one or more of the plurality of locations of interest.
20. The position tracking system according to claim 19, wherein the plurality of locations of interest include the location of at least one of: a wireless beacon; a specified product; an aisle; an entry point to an area (e.g. a store entrance, aisle entrance, cart park and/or the like); an area perimeter (e.g. a security perimeter); a shelf; a wall; a checkout area; another cart; and/or a staff member.
21. The position tracking system according to any of claims 16 to 20, further comprising means for tracking a route taken by the cart.
22. A method of deriving positional information for a cart for carrying items, the method comprising: acquiring, at a wheel of the cart, information representing a rotational distance travelled by the wheel; and outputting information acquired by said acquiring step for estimating a change in a position of the cart based on said information acquired by said acquiring step.
23. A method performed by a position tracking system for deriving information relating to a location of a cart according to any of claims 1 to 4, the method comprising: communicating with the cart, to obtain information relating to an estimated change in a position of the cart; and identifying, based on said information obtained from the cart, at least one location at which the cart was located.
24. A computer implementable instructions product comprising computer implementable instructions for causing a programmable computer device to perform the method of claim 22 or 23.