FIELD OF THE INVENTION

The present invention relates to a road condition sensor, and more specifically means for determining the condition of a road surface and identifying any contaminants present thereon.

BACKGROUND OF THE INVENTION

The contact between the tire of a vehicle and the surface upon which it is acting is vital to the performance of the vehicle, as well as the safety of operation of the vehicle. This can be affected by many factors, such as the condition of the tire, the surface upon which the tire acts, as well as the presence of any contaminants on the road (for example, water, snow, chemicals, biological matter such as leaves, etc.), as well as the thickness of the contaminant.

Historically, the presence of any contaminant on the surface has been determined by the operator of the vehicle seeing the contaminant, and then making adjustments to the operation of the vehicle accordingly. However, there still exists a need to more accurately and reliably detect a wide range of road contaminants, and how this effects the operation of the vehicle, particularly in self-driving vehicles.

By accurately determining the presence of road contaminants, and how they affect the operation of a vehicle, it is possible to modify the operation of a vehicle so as improve its efficiency.

Several methods have been previously proposed as a way of determining road conditions.

A limited number of photodiodes, which convert light at specific wavelengths into an electrical current, have been used to determine the presence of water and ice. Such systems use photodiodes configured to measure light that is specifically absorbed by water and ice. In addition, lookup tables containing a list of reference data have been considered, which compare a set of input data from the photodiodes representing the road condition to a number of known conditions, and then using the best match to give an estimate of what the road condition is like, i.e. whether there is water and/or ice on the roads. These systems are not very robust, and can also require a lot of storage space for the reference data, which limits its real time use.

Time-varying filters have also been used with a single photodiode. This method uses a time-varying filter (e.g. a micro-mirror device or fabry perot filter), and commonly a photodiode to measure the light intensity at only one point. Whilst such an approach is low cost, each wavelength measured is taken from different areas of the road when the vehicle is moving, as the measurements are taken over different times, thereby limiting its usefulness.

SUMMARY OF THE INVENTION

In a first aspect, according to the present invention, a system for mounting on a vehicle to detect a contaminant or condition of a surface upon which the vehicle is travelling is provided as claimed in claim 1. The system comprises a detection module comprising; a sensor having a field of view of the surface, in use, the sensor comprising; an array of photodetectors; and a spatial-variable wavelength filter configured to limit the wavelength of light that is incident on each photodetector of the array of photodetectors. The detection module further comprises an optical engine configured to collimate light incident on the sensor; and the system further comprises a light source configured to, in use, provide illumination of the field of view of the sensor. Such a system may be able to detect far more accurately and robustly a contaminant or condition of a surface.

By providing a spatial-variable wavelength filter with a corresponding array of photodetectors, each photodetector may detect the intensity of only a specific wavelength or range of wavelengths of the reflected irradiated light that is allowed to pass though part of the wavelength filter adjacent to and corresponding to each photodetector. An irradiation spectrum may then be built up, and may then be analysed so as to provide information as to the presence of a contaminant or the condition of the road. To this point, the system may further comprise a controller, the controller configured to receive irradiation spectrum data from the sensor and analyse the irradiation spectrum data so as to determine one or more characteristics of a contaminant or condition of the surface.

The controller may be further configured to identify the contaminant and/or the thickness of the contaminant, and the controller may additionally or alternatively be configured to estimate the tire to surface friction coefficient based on the determining the presence of a contaminant or condition of the surface. The controller also or alternatively may be further configured to identify the chemical composition of the contaminant.

The array of photodetectors of the system above may comprises at least 5 photodetectors. By providing such numerous photodetectors, a more accurate irradiation spectrum may be populated that contains more information about the condition of the surface or the contaminants on the surface.

The wavelength of light able to pass through the spatial-variable wavelength filter may vary linearly across the filter, or discontinuously across the filter. A linear variation may be easier to produce and give a broader indication of the absorption spectrum, but a discontinuously varying filter allows for specific wavelengths of interest to be magnified. This is particularly useful in situations where specific contaminants have a characteristic irradiation spectrum regarding only a few wavelengths.

The wavelength of light able to pass through the spatial-variable wavelength filter may be within the Short Range Infra-Red band.

To reduce the amount of light reflecting directly off the surface/contaminant and into the sensor, preferably the light source and the detection module are angled relative to one another and/or to the surface. For example the light source and the detection module may be angled by 10° to 30° to one another. Additionally and/or alternatively, the light source and the detection module may be provided at an angle of less than 90° with respect to the surface. The optical engine may comprise at least one aperture configured to restrict the angle at which light is incident on the sensor. Additionally or alternatively, the optical engine may comprise a focussing lens and/or a collimating lens. Advantageously, apertures prevent dirt from passing through the optical engine to the sensor. A focussing lens or lenses may increase the intensity of light incident on the sensor, and a collimating lens or lenses ensures that the light incident on the sensor is within an angle that may be measured by the sensor. A focussing lens and a collimating lens may be provided as a pair.

The spatial-variable wavelength filter may vary progressively. By providing such a variable wavelength filter, it is possible to perform a form of spectral detection (i.e. measuring a spectrum, rather than discrete points within a spectrum) sufficient to allow chemical identification not necessarily limited to specific substances.

In a second aspect, there is provided a vehicle. The vehicle comprises the system as described above mounted on the vehicle, such that the sensor is configured to have a field of view of a surface upon which the vehicle is travelling, in use, and the light source configured to, in use, provide illumination of the field of view of the sensor.

In a third aspect, there is provided a method. The method comprises mounting the system as described above to a vehicle such that the sensor is provided with a field of view of the surface upon which the vehicle is configured to travel, and the light source illuminates the field of view of the surface.

In a fourth aspect, a method for detecting a contaminant or condition of a surface upon which a vehicle is travelling is provided. The method comprises illuminating a field of view of the surface; collimating light emitted from the field of view of the surface; measuring the irradiation spectrum of light emitted by the surface using a sensor comprising an array of photodetectors and a spatial-variable wavelength filter configured to limit the wavelength of light that is incident on each photodetector of the array of photodetectors; and determining the presence of a contaminant or condition of the surface. ln the method the step of determining the presence of a contaminant or conditions of the surface may further comprise identifying the contaminant and/or the thickness of the contaminant. Additionally or alternatively, the method may further comprise the step of estimating the tire to surface friction coefficient based on the determining the presence of a contaminant or condition of the surface.

Additionally or alternatively, the step of determining the presence of a contaminant or conditions of the surface may further comprise identifying the chemical composition of the contaminant. Additionally or alternatively, the spatial-variable wavelength filter may vary progressively.

BRIEF DESCRIPTION OF THE DRAWINGS

Preferred embodiments will now be described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 shows a Road Condition Sensor as described herein.

Figure 2 shows four example sensors showing different photodetector arrangements and spatial-variable filters that may be utilised therewith.

Figure 3 shows an optical engine that may be used in the road condition sensor as described herein.

Figure 4 shows an example relative field of view of the sensor and an illuminated area from the light source.

DETAILED DESCRIPTION OF THE INVENTION

Whilst the invention will be described herein with reference to road vehicles, such as cars or trucks and their tires on operation on a road surface, it is envisaged that this invention may be equally applicable to any vehicle in determining the presence of any contaminant on the surface on which it operates. For example, it is envisaged that the present invention may be equally applicable to trains and/or trams, and determining the presence of any contaminants on the tracks upon which they operate. Equally, the present invention may also be utilised in airports, on vehicles and/or aircraft on runways. The present invention may also be utilised in both advanced driver-assistance systems (ADAS), aiding a vehicle driver whilst driving, and self-driving cars.

By improving the detection of the condition of the road, it is possible to greatly increase the situational awareness of the vehicle and the means for driving the vehicle, taking into account the exact road conditions, thereby significantly increase the road safety. Accurate and local information for driver and car onboard systems offers the opportunity to mitigate risks preventively and effectively.

A road condition sensor 100 (RCS) in accordance with the present invention is shown in Figure 1. Such an RCS 100 may be utilised to determine characteristics of what is (or indeed what is not) on the road. The RCS 100 may be able to detect various properties of the road surface, including the material of the road surface itself, as well as the presence of any contaminants on the surface of the road. For example, the RCS 100 may detect the presence of contaminants such as water, snow, chemicals and their chemical composition, biological matter such as leaves. The RCS 100 may also detect how much of the contaminant is present (for example, the extent of the contaminant and the thickness of the layer), as well as the effect that any such contaminant has on the road qualities. For example, how the specific contaminant/amount of contaminant effects the tire to road friction, or at what point the contaminant would freeze. Such an RCS 100, in use, is to be mounted onto a vehicle (e.g. a car or a truck). In use, the RCS 100 may be mounted under the vehicle. The detection of other contaminants is also envisaged, as are other mounting positions for the RCS 100.

The RCS 100 of the present invention comprises a detection system (comprising a sensor 10 and an optical engine 3) and a light source 4. From the light source 4, light 5 is emitted, which is incident upon surface 6. The presence of a contaminant 7 on the surface 6 can reflect and/or absorb certain wavelengths of the light from the light source 4. By measuring the reflectance of the road and the absorption by the contaminant, it is possible to determine relevant information about the contaminant characteristics. The sensor 10 comprises at least one photodetector, for example in an array of photodetectors 1, and a filter restricting the wavelength of light that reaches each photodetector of the array of photodetectors, such as a spatial-variable wavelength filter 2. An increasing number of photodetectors will be able to specifically detect more wavelengths, and thereby provide increasing robustness and the ability to reliably detect more parameters, such as chemical concentrations. Preferably, the array of photodetectors comprises at least 6 photodetectors, or even more preferably at least 15 photodetectors. The array of photodetectors may comprise between 15 and 40 photodetectors, or may comprise at least 40 photodetectors. In one particular example it is envisaged that an array of 128 photodetectors is used in the sensor 10.

Four example sensors 11, 12, 13, 14 showing different photodetector arrangements are shown in Figure 2. The spatial-variable wavelength filters 201, 202, 203 and 204 are represented by the gradients in wavelengths that are allowed to pass through the sensor, as seen in each sensor. As would be appreciated by the person skilled in the art, the shading represented by the gradients in Figure 2 are only representative of how the filter changes so as to allow different wavelengths to pass through the filters at different spatial points, and thereby the gradients do not represent the true colours let through.

A wide range of wavelengths may be analysed by the RCS 100, or alternatively a sub-section of wavelengths can be selected for analysis depending on the contaminants that are to be detected. For example, sensor may be configured to detect the Short Wave Infra-Red (SWIR) band, between 900 - 1700 nm. The characteristic absorption spectra of water in different states has wavelengths in this range.

However, as would be appreciated by the skilled person, the range can be expanded or adjusted in order to find other contaminants or qualities. For example, a water with various concentrations of a certain de-icing salt used in airports may absorb different wavelengths than contaminants causing a slippery road surface for passenger cars and trucks, e.g. rainwater and ice (another possible use). The spectra to be analysed may therefore customised, depending on the desired use of the RCS. By providing a spatial-variable wavelength filter with a corresponding array of photodetectors, each photodetector may detect the intensity of only a specific wavelength or range of wavelengths of the irradiated light that is allowed to pass though part of the wavelength filter adjacent to and corresponding to each photodetector.

As may be seen in Figure 2, sensor 11 shows an array of photodetectors 101 laid out horizontally, and the filter wavelengths of the spatial-variable wavelength filter 201 varying linearly from 1700nm to 900nm.

In sensor 12, the array of photodetectors 102 are similarly laid out horizontally, but the filter wavelengths of the corresponding spatial-variable wavelength filter 202 vary non-linearly. For example, a discontinuity in the wavelengths of the filter may be found near the middle of the filter at point 302, with the red-green range being more drawn-out than the blue-purple range. For example, at point 302, the wavelengths of 1100nm to 1400nm may be omitted. This results in the ranges of 900 to 1100 nm and 1400nm to 1700nm being more drawn out. Of course, any subsection of wavelengths from a larger total wavelength may be omitted. Equally, there may be more than one discontinuity point in the spatial-variable wavelength filter.

Advantageously, by providing such a discontinuous filter it is possible to dedicate more photodetectors to detecting wavelengths of interest. To this point, it is possible to dispense with unused wavelengths, and receive more data certain reflected wavelengths that can be analysed to give more information about, and show different characteristics of the detected contaminant.

Sensor 13 comprises an array of photodetectors 103 laid out in two dimensions (x,y), whereas the wavelengths allowed to pass by the spatial-variable wavelength filter 203 vary linearly in only one dimension (x). Similarly, photodetector array 104 of sensor 14 is laid out in two dimensions (x,y), but the wavelength allowed to pass through the spatial-variable wavelength filter 204 varies linearly in two dimensions (x, y), for example from the upper left corner to the lower right corner. It is also envisaged that a spatial-variable wavelength filter may vary discontinuously in two dimensions, with a suitable two-dimensioned array of photodetectors. By providing such a two-dimensional array of photodetectors, and a corresponding spatial- variable wavelength filter, more data may be collected from each sensor. With such sensors, it is envisaged that imaging optics could also be used, allowing such sensors to sense different wavelengths of light for different parts of the road.

As would be appreciated, whilst Figure 2 shows filters that allow light with a wavelength of between 900nm and 1700nm to pass through at different points, these filters could be provided so as to allow lights of other wavelengths through, or to omit any wavelengths.

The provision of a relatively large number of photodetectors and a matching spatial- variable wavelength filter allows for detection of many wavelengths. By seeing such a fine spectrum of light, the RCS 100 of the present invention may not only detect the presence and level of water and ice, but also determine contaminants in fine detail. For example, details such as presence of salt in the water, as well complex scenarios such as water on top of ice and wet leaves, and the presence of other chemicals (such as oils or anti-icing chemicals etc.) can be detected. The detection of a multitude of different contaminants and/or a combination of contaminants which all have different effects on the secondary characteristics of the road (such as friction, freezing point etc.) allows for a much finer degree of accuracy in determining the secondary characteristics caused by the contaminants.

The use of a spatial-variable wavelength filter instead of a temporal-variable wavelength filter is preferable. The spatial-variable wavelength filter allows a multitude of photodetectors to read a multitude of wavelengths at the same time, while a temporal-variable wavelength filter allows only one photodetector to read a multitude of wavelengths, over a time period (one at a time). By utilising a spatial- variable wavelength filter, it allows for faster sensor readings, which is particularly beneficial when the RCS 100 is mounted on a fast-moving vehicle. Additionally, such an arrangement does not require any moving parts, thereby improving reliability.

The present invention utilises an optical engine 3 (for example, as seen in Figure 3) operatively connected to the sensor 10 to collimate light entering the RCS 100. Such an optical engine 3 may ensure that any incident light thereon fulfils any requirements of the array of photodetectors 1 and spatial-variable wavelength filter 2 of the sensor 10. For example, the sensor 10 may only be able to accurately measure rays of light within a certain angle of incidence. The optical engine 3 may therefore provide a collimating tube and/or one or more collimating lenses.

Additionally and/or alternatively, the amount of light noise may be reduced by providing an optical engine 3 with an inner surface with low reflectivity, such that light cannot errantly reflect around the inside the optical engine 3. The optical engine 3 may also provide cooling of the sensor 10, and/or contain bandpass, lowpass and/or highpass filters to remove wavelengths outside the range provided by the spatial-variable wavelength filter 2 attached to the array of photodetectors 1. The optical engine 3 may also increase the light intensity that reaches the photodetectors, for example by using one or more focussing lenses.

An optical engine 3 for use with the present invention may be seen in Figure 3. The optical engine may be a collimating tube. The optical engine may consist of an outer tube 31, which protects the components within and absorbs any light inside, by virtue of the inner surface of the tube having very low reflection. The outer tube 31 may comprise at least one means for restricting light from entering the optical engine 3, such as at least one slit and/or aperture. In this way, it is envisaged that one means for restricting light may be present, or equally that two or more means for restricting light may be present in the optical engines. A focussing lens and collimating lens pair may be additionally positioned in front of and/or behind the apertures.

For example, in the embodiment of Figure 3, three apertures 32, 33, 34 are present. The apertures are thin, round plates, that prevent light entering the optical engine 3 at an in an angle that is too great for the spatial-variable wavelength filter to properly operate. The apertures may also stop dirt from reaching the filter 10. Whilst, in the embodiment of Figure 3, the outer tube 31 and apertures 32, 33, 34 appear round, they may also be any shape. For example, the outer tube and apertures may be elliptical, rectangular, or rectangular with rounded corners to provide a more even intensity and angle of incidence of light onto the sensor. Figure 3 shows the collimating properties of the optical engine 3. Ray 36 may enter the optical engine 3 perpendicularly to the surface of the sensor 10, and ray 37 is at an angle a from ray 36, in two dimensions. Such an angle a may be the largest angle of incident light that may be measured by the photodetectors 1 and spatial- variable wavelength filter 2 of the sensor 10, and the apertures may prevent light at any angle too great for the sensor 10 from reaching the spatial-variable wavelength filter.

The RCS 100 of the present invention does not rely on imaging optics. In this way, the RCS 100 does not create an image of the field of view of the photodetectors. In this case, it is desirable that light from one point of the road surface is not incident at only one point on the area of the photodetectors. Rather, and advantageously, each photodetector of the array of photodetectors of the present invention receive light from the full field-of-view of the sensor. In this way, the RCS 100 is more robust to small objects of a different colour, which could otherwise distort the wavelength spectrum.

Advantageously, by using apertures as opposed to a slit, it is possible to allow more light to reach the photodetectors. This is particularly beneficial when high speed measurements are required, e.g. when the sensor is attached to a vehicle driving at high speed and its data will be used for vehicle dynamics and manoeuvring.

The light source 4, as can be seen in Figure 1, emits a beam of light 5. The light source 4 may be chosen so as to emit light of the wavelengths that are characteristically absorbed by the contaminant to be detected. In this way, the wavelengths emitted by the light source 4 may be the same wavelengths that the sensing 10 (such as the array of photodetectors 1 and spatial-variable wavelength filter 2) are optimised for.

The light source 4 may be a wide-wavelength light source, such as black-body emitters (e.g. halogen lamps). Additionally and/or alternatively, the light source may also comprise one or more narrow-wavelength light sources, such as LEDs. For example, the light source 4 may comprise one wide-wavelength light source, a multitude of wide-wavelength light sources, and a multitude of narrow-wavelength light sources. The light source 4 may comprise a halogen light bulb. Such a bulb radiates a lot of light in the wavelengths 900-1700 nm, which has been found to be particularly useful in detecting contaminants. Additionally, such a bulb is cheap, has a long lifespan, is commonly available in voltage range commonly used in the automotive industry (for example, 12-24 V). As would be appreciated by the skilled person, another black-body emitter with similar characteristics may be utilised.

It is envisaged that a multitude of LEDs over a range of wavelengths could equally provide similar benefits. Furthermore, whilst in the embodiment of Figure 1 a single light source 4 is shown, it is also envisaged that a multitude of light sources may be provided around the optical engine 3. In such a case, smaller light sources may be provided, which may improve the lifespan of the light source 4. The light source 4 is orientated such that it provides a beam of light incident on a surface 6 such that the field of view of the optical engine 3 is illuminated, and thereby the field of view of the sensor 10. On the surface 6, there may be a contaminant 7. To this point, the RCS 100 may be able to detect the presence of any contaminant 7 on the road, as well as the amount of contaminant, for example the depth of the contaminant. The RCS 100 may then determine the contaminant’s effect of other road properties, such as the freezing point and friction in real time, whilst mounted on moving vehicles, such as trucks, busses and cars. Whilst the most common types of contaminants are likely to be water, ice and snow, the RCS 100 may be able to detect other contaminants, if necessary.

The RCS 100 may therefore be further able to detect the concentration and level of specific chemicals on a surface, for example anti-icing chemicals on the runway of airports.

The detection system (consisting of the sensor 10 and the optical engine 3) and the light source 4 should be mounted together such that the light source 4 lights up an illuminated area 51 on the surface 6 such that the sensor tube field 52 of view is illuminated as uniformly as possible. It is therefore beneficial that the light source 4 and the detection system are mounted with as short a distance and as low an angle between them as possible. The angle between the detection system and the light source 4 may be seen in Figure 1 as an angle of rotation around the z axis in Figure 1. By providing a low angle between the detection device and the light source 4, the sensor field of view may remain centred in the middle of the illuminated area 51 on the surface 6, when the sensor-to-ground distance changes.

However, a low angle of incidence may also result in light reflections from the road surface 6 or contaminant 7 reaching the sensor 10. As the sensor 10 measures the absorption of the road surface 6 and/or contaminant 7, any specular light reflection is undesirable. Therefore, in order for only diffuse light to reach the optical engine tube 31, and thereby the sensor 10, both parts are rotated relative to each other, and relative to the road surface 6. For example, in figure 1, the light source 4 and the detection system have been rotated relative to one another around the z-axis. Additionally, or alternatively, the light source 4 and the detection system may be rotated together, relative to the road surface. Again taking figure 1 as an example, the light source 4 and the detection system may be rotated around the x-axis, so as to not be located in a plane perpendicular to the road surface and reduce the amount of light that reflects directly from the light source into the detection system. Such an arrangement may create an illuminated area 51 and field of view 52 as seen in Figure 4. Preferably, the light source 4 is mounted at an angle of 5°-30°, relative to the sensor 10, and/or preferably the light source 4 and the detection system may be rotated together relative to a plane defined by the road surface so as to be at an angle of less than 90° to the road surface, or more specifically an angle of 5-50°.

Absorption spectra data collected by the sensor 10 of the RCS 100 may be provided to a controller (not shown) and processed. The data may be processed using data-driven algorithms (i.e. Machine Learning) that is able to classify the contaminant, its composition, and find its layer thickness based upon the absorption spectra received by the sensor 10. In order to produce the algorithm, a vast amount of reference data is collected, and then the algorithm is trained using this data, adjusting parameters depending on the data input and reference output. By using such a data-driven algorithm, it is possible to produce a system that is robust to new and unknown scenarios, and unknown contaminants. Further, data-driven algorithms may also take into consideration a vast amount of different scenarios without having to store all the scenarios on-board or handle them in real-time. During processing, prior to and after machine learning, various filters (e.g. Bayesian filters and smoothening filters) may be used.

Advantageously, the RCS 100 of the present invention may be able compensate for an altered (skewed) spectrum that may be present, when the field of view of the RCS is in strong sunlight. The temperature of the ambient light may be estimated using one or more additional photodetectors in the visible wavelength range (e.g. red, green, blue), and using such data, the absorption spectrum of the road/contaminant and detected by the sensor 10 is shifted before it is input into the machine learning algorithm for determining the presence and thickness of contaminant.

Additional sensors may also be used to measure ambient temperature and humidity, road surface temperature, vibrations, distance between the sensor and road, angle of inclination. Data from the car may also be collected through the Controller Area Network (CAN) bus interface, such as throttle and tire air pressure. This data is used in the algorithm as part of either or both of analytical models (e.g. road weather models, freezing point models, friction models) and as direct input to the machine learning model. The additional sensors will give additional information about the environment, and will make the algorithm more robust. E.g. if the road temperature is 20°C, it is unlikely that the road will be icy, and when the humidity is high, there is a higher possibility that the road surface is wet. All of this information can be used in determining the estimated tire to surface friction, or other qualities such as the chemical composition of any contaminants on the road.

The accurate detection of the tire to surface friction allows for operation of the vehicle to be optimised, therefore improving the efficiency of operation of the vehicle, reducing energy wastage, and thereby reducing fuel/energy costs as well as potentially dangerous emissions. Advantageously, in situations where chemicals/contaminants have been purposely added onto the surface (for example, where chemicals are applied to the surface in cold temperatures to prevent freezing), the RCS may monitor the depth and chemical composition of the contaminant, and thereby determine when a suitable amount has been applied to the surface to provide its desired effect. It therefore may be possible to prevent excessive amounts being applied to the road surface which is not only wasteful, but can also have a negative effect on the surrounding environment, if such chemicals are subsequently washed off the road into the environment.

Although the invention has been described in terms of preferred embodiments as set forth above, it should be understood that these embodiments are illustrative only and that the claims are not limited to those embodiments. Those skilled in the art will be able to make modifications and alternatives in view of the disclosure which are contemplated as falling within the scope of the appended claims.