Field of the Invention The invention relates generally to the measurement of the velocity of a human in motion relative to the air, most usefully on a bicycle. In the absence of wind, the invention relates to measurements of the actual speed of motion, or, in the presence of wind, the relative effect of the wind on the effort of the human. The invention also relates to the integration of the measurement of the velocity of a human in motion with numerous other measurements that are conventionally made, and obtaining a more complete picture of the athletic performance of the human. The invention also relates to obtaining a data log of the measurements made in real time and the capability of storing, processing and reviewing the data. Further, the invention relates generally to methods of calibration which give accurate measurements corrected for any local aerodynamic properties.

Background of the Invention The measurement of speed on bicycles is very well known in the art. The first cyclometers consisted of a purely mechanical device, where a pin attached to a spoke would mechanically engage with a stationary counter device and simply count the number of revolutions. With advent of microprocessors, the use of a magnet delivering impulses to a receiver connected to a display on the handlebars permitted the display of speed in real time, as well as performing simple computations of distance derived from the wheel circumference, average, maximum, minimum speed and the like. Such devices 10 have come into general use. There is currently a proliferation of improvements with increasing microprocessor power, such as the integration with heart rate measurements as exemplified by the art described in US Patent 5, 810, 722"Method and Device for Determining Threshold Values for Energy and Metabolism"by Heikkila and products commercialized by Polar Electro Oy, measurement of force delivered to the chain as in US Patent 6,356, 848"Method and Apparatus for Measuring Power Output and For Measuring Tension and Vibrational Frequency of a Elongate Flexible Member"by Cote et al, also commercialized by Polar Electro Oy, integration with pedal cadence, altimeter and temperature measurements and heart rate as commercialized by Ciclosport AG.

20 There has been a parallel proliferation of technology for the sport of running, starting with the pedometer which was manually clicked for each pace, and had to be calibrated for individual stride, to current generation inertial devices which transmit a signal to a wrist-worn display, and require no individual calibration, as exemplified by products of the Nike corporation; and Global Positioning System devices which measure distance and speed with a high degree of accuracy, and also function through a watch display, the watch having all of the other features of a sports chronometer, including memory for lap and split times, and capability of numerous computations such as average speed, maximum speed etc, as exemplified by a products from the Timex Corporation.

Even newer developments integrate a heart monitor and associated computations into the 30 same functional system with a wireless connection to a wrist watch, and the entire data set may be downloaded on to a computer.

The foregoing shows that there is a burgeoning proliferation of advanced technologies for use with monitoring and evaluating athletic performance and increasing demand for more complete and accurate information concerning the athletic performance and for evaluating the effect of training schedules.

Wind velocity measuring devices, also called anemometers, have a long history in weather observation as part of recording of wind velocity and direction. The traditional device in usage in use since it's invention in 1846 by John Thomas Romney, consists of 10 four hemispherical cups mounted at the end of a horizonatally rotatable cruciform support, and measurement of wind velocity from the speed of rotation of the assembly.

More recent improvements for use in measuring air flow in ducts and ventilating systems, needed by heating and air-conditioning engineers, are more rugged and portable, and consist of a rotatable vane in a rugged housing and connected to a microprocessor controlled display system. The display can show results in a variety of units and show air velocity, or, as is sometimes needed, air volume flow. Such devices, as commercialized by the Extech Instrument Corporation, for example, are combined with humidity and temperature measurements. A further improvement in miniaturization and ruggedness, needed by said engineers, functions on the hot wire principal. The convective cooling of 20 the hot wire can be directly related to rate of air flow. The hot wire anemometer only require accurate thermometry, easily attainable with current thermistors, and has the advantage of having no moving parts. The vane and hot wire anemometer devices also have the capability of logging data, including temperature and humidity measurements, as well as air flow, permitting the review of the performance of a heating/airconditioning system over a period of time.

The foregoing considerations show that there are numerous possible solutions to the problems of air flow measurement or movement relative to the air, which are capable of miniaturization and combination with data processing, microprocessor utilization, and 30 electronic display means according to currently known art.

Another area where air velocity measurements are critical is in aeronautics, where navigation, flight planning, fuel usage considerations and numerous other factors depend critically on the knowledge of air flow, or, in other words, motion relative to the air, known as airspeed. While speed relative to the earth's surface, or ground speed, is important in calculation of such things as expected time of arrival at a fixed destination, aircraft pilots need to know what their air speed is for considerations of navigation, power utilization and fuel useage. For most of the history of aviation, the airspeed has been measured with a Pitot tube. The original Pitot tube was invented by Henri Pitot in 1732 and made practical for liquid flow measurments by Henry Darcy in 1856. The ram 10 effect of the oncoming air will result in an increase in barometric pressure relative to an aperture that is not exposed to the flow. As the air speed increases, so does the ram effect, so that difference of the barometric pressure at the two locations is a very reliable measure of air speed. Examples of early devices for using such principles for determining aircraft speed are to be found in US Patent 1,146, 202, "Anemometer"inventor A. Ogilvie and US Patent 1, 290, 875, "Air Speed Indicator for Aeroplanes", inventors A. T. Baldwin and S. L. Christenson. This system is still the standard for use in air speed measurement in aeronautics. Note that Ogilvie depicted the Pitot tube mounted on a support vertically well above the body of the plane, so as to be clear of any spurious aerodynamic effects of the propeller and turbulence from the body. Another early device described in US Patent 20 1,436, 575, "Air Speed Indicator", inventor C. R. Colt, utilized a small air-driven propeller driving an electrical generator whose signal was measured in an ammeter, which, while it did not become established as the preferred method of airspeed measurement, it can be considered as the precursor of current generaton of anemometers mentioned above. It is also noteworthy that on current generation aircraft, the pitot tube is usually mounted on a wing to be well clear of spurious aerodynamic effects of"propeller wash"as well as clear of any control surfaces on the wing.

Following are examples of the prior art for determination of speed, performance and power utilization in sports related activities. US Patent 4,352, 063"Self-calibrating 30 Speedometer/Odometer", inventers P. W. Jone and D. W. Purcell, describes the use of the magnetic interaction of individual spokes to provide electrical impulses for accurate speed determination; this invention also provides means for calibration of spoke number to determine accurate distance travel. US Patent 4,156, 190"Electronic Bicycle Odometer and Speedometer", inventors B. C. Chittenden and R. E. Hay describes a method of bicycle speed determination with no moving parts by means of a light beam that is interrupted by the passage of the wheel spokes. Cadence, or number of pedal rotations per minute, measurements are also commonplace in determining performance and performance goals for a bicyclist, and an example of such is a bicycle-mounted accelerometer in US Patent 4,526, 036"Cadence Meter"by T. R. Morrison. Doppler effect of reflected sound waves is a known method of speed determination, as exemplified by US Patent 4,914, 638 10"Doppler Effect Speedometer"by R. E. Miller and references therein. This latter invention is combined with speech synthesizer output so as to be particularly useful in sports such as running, where the visual observation of a display is inconvenient. US Patent 4, 780, 864"Combination wristwatch and bicycle computer", inventor Houlihan, describes the use of a wrist watch display mounted on the handle bars of a bicycle measuring speed and cadence via magnets mounted on a spoke and on a sprocket and sensors positioned to detect the passing of the magnets. US Patent 5,003, 820"Electronic Speedometer for Snow Skis", inventor E. Dittbrenner teaches the use of a device based on a thermistor and a thermocouple for speed measurement in skiing or hang gliding based on the hot wire principle. Other art for determination of speed in skiing is also described therein. There is 20 no description of how one would calibrate the device, and no mention of how to deal with local turbulence conditions. US Patent 6,539, 336"Sport Monitoring System for Determining Airtime, Speed, Power Absorbed, and other Factors such as Drop Distance" by C. A. Vock et al. and other art therein describes a variety of sensors for all manner of sports-related movements and activities. They describe the use of"a truth sensor, such as police radar gun, or a simple wind anemometer"to verify a specific computation. They do not elaborate on how this would be achieved in practice, and again, do not consider the local air flow conditions for such an anemometer. The above-mentioned US Patent 6,356, 848 by Cote et all ingeniously measures power output by the bicyclist, by means of determination of speed and tension in the chain from the vibration frequency of the chain 30 of the bicycle. This has the disadvantage, however, of not being able to determine wind effects directly and also requires critical adjustments and alignments of physical components, such as the sensors that monitor moving parts, to obtain satisfactory measurements. The power output may be integrated over a period of time to determine calorie consumption. Another method for estimating energy consumption is described in US Patent 6,537, 227"Method and Equipment for Human-Related Measuring", inventors H. Kinnunen and S. Nissila. This method uses a combination of heart rate and one or more of a number of personalized parameters. However, the personalized parameters are tedious to determine accurately, depend critically on the individual state of fitness, and can vary unpredictably due to transient changes in the physical health of the individual, diurnal variations and the like. Similarly, the software provided by Ciclo Sport with their 10 bicycle computer system provide calculated values for power output without taking into account any variables other than biometric data and the groundspeed. As will be shown, the present invention permits the accurate determination of power output and energy consumption directly from objectively measurable parameters, with no simplifying assumptions.

The aerodynamic characteristics are crucial for high performance bicycling. Skin tight clothing and pear-drop-shaped helmets are worn. The lowered stance with drop- handlebars is used to lower the frontal resistance, and wheels with aerodynamically designed spokes for the wheels are examples. Also, drafting of one cyclist behind a group 20 of cyclists is a well-known method of saving energy by reducing the cycle speed relative to the air in the slip stream. Thus, it is eminently clear that aerodynamics is considered of critical importance in performance bicycling. Bearing this fact in mind, any significant wind will also affect performance, in direct relation to the speed of the wind. However, no device has been adapted for using in airspeed measurements for bicyclists. Further, elaborate algorithms are provided with current bicycle computer technology for determining calorie consumption. These take into account rolling resistance dependent on the weight of the bicyclist and tire characteristics, and chain tension, heart rate and work to achieve changes of elevation. However, to date, no practical method has been achieved for measuring and computing the effects of wind resistance or wind velocity.

30 Methods have been described for amateur construction of an anemometer, and use of a bicycle speedometer for calibration thereof (see http://www. astro. uni- bonn. de/-kbagschi/anemoe. shtml). As stated on that web site"I chose a windless (as far as this is possible) weekend day and attached the anemometer to a long piece of timber, about one and a half meters in front of my bicycle. Then I rode up and down our (very quiet) street comparing the displays of the two odometers (bicycle and anemometer)."In that case, the anemometer was mounted far forward of the bicycle on a pole, in order to eliminate aerodynamic disturbances due to the cyclist and rotating wheels, and a windless day was recommended as suitable conditions for calibration. Another example"So now 10 that all things are ready I only have to wait for a zero wind day to get out with my anemometer placed over the bicycle, in a height, so that turbulence from the bicycle and me don't influence the result. " (see http://home. worldonline. dk/bhm/Anemometere. htm) This would also tend to discourage one from directly mounting the anemometer in a more practical way, and suggests the impossibility of performing such a calibration in the presence of any significant wind. It is also very difficult and inconvenient to wait for and select rare weather conditions which are completely windless. The current invention shows surprisingly simple methods for circumventing this problem. Furthermore, the foregoing considerations are for calibration of an anemometer, and are not suggestive of providing an instrument for measurement of airspeed in human-powered motion.

20 It is the purpose of the current invention to describe practical ways of mounting an anemometer on a bicycle, or other human-powered vehicle or craft, at a location which is physically convenient, to utilize data obtained from motion of the bicycle relative to the air, to show methods of calibration which compensate for local air disturbances caused by the bicycle, and thus obtain surprisingly accurate, reliable and reproducible methods for determining such motion, and the integration of such data with data more conventionally obtained with bicycle computers, and thus giving a more complete determination of performance. The current invention also permits the tracking of performance over journeys and determination of effects of wind over different parts of a 30 journey, together with other physical parameters. Further advantages over the present art will be described for more accurate determination of power delivered or calorie

consumption, taking accurately into account aerodynamic as well as other more conventional factors.

Definitions The following terms are used in this document, following conventional uses. Since the conventional uses are in different fields which overlap herein, for clarity, they are used as follows.

"Anemometer"is a stationary device for measuring the windspeed, such as weather conditions or airflow in ventilation systems.

"Windspeed"is the airflow usually measured in velocity units, such as miles per hour or knots, and is conventionally used as the units measured by an anemometer.

"Airspeed"is nominally the same as"Windspeed"but is used in aeronautics for movement relative to the air. It can be measured by anemometer-like devices but usually different in practice because of the different speed ranges involved in flight compared with weather. When there is a net windspeed component opposing the direction of motion, then the airspeed is the groundspeed plus the component of the vector of the windspeed in that opposing direction. Conversely, when there is a net windspeed in the same direction as the direction of motion, then the airspeed is the groundspeed minus the component of the vector of the windspeed in the direction of motion.

"Airspeed indicator"is a device which measures and displays the airspeed, and functions analogously to an anemometer.

"Apparent Airspeed"or"Aas"is the uncorrected data measured directly by an airspeed indicator.

"True Airspeed"or"Tas"is the corrected value for the airspeed, correct by means disclosed in the embodiments disclosed herein.

"Groundspeed"is the speed relative to a fixed surface, such as the road surface in the case of road vehicles, or the surface of the earth below, in the case of aeronautics.

It will be clear from the following that the clarity of these definitions is crucial, as the present invention, for the first time, brings considerations and capabilities into the field of human-powered propulsion, that were hitherto the exclusive reserve of the field of aeronautics.

Summary of the Invention The present invention consists of methods and devices for the determination of speed relative to the air for human powered transportation, for providing improved methods of computing performance and power consumption, and for utilizing such 10 methods and devices for provision of data compared with performance goals, for improved methods of determining calorie consumption, and for designing strategies for optimizing aerodynamic performance. The invention also provides for improved methods for calibration of such devices, as well as the use of combination of airspeed together with more conventional sport-related data, both by continuous data logging and by real time readout. The present invention also provides for determination of the wind component and its effects, in a journey, optionally as part of logging other kinds of sport- related data.

Brief Description of the Drawings 20 Figure 1 represents a hybrid-type bicycle and airflow patterns encountered.

Figure 2 represents a hybrid-type bicycle showing mounting positions for airspeed measurement device.

Figure 3 represents a road bicycle and airflow patterns encountered.

Figure 4 represents a road bicycle showing mounting positions for airspeed measurement device.

Figure 5 shows the controller unit of an anemometer.

Figure 6 shows an anemometer.

Figure 7a shows the top view of the mounting of the anemometer control unit Figure 7b shows the side view of the mounting of the anemometer control unit 30 Figure 8 is a cross section showing the mounting of an anemometer on handle bars.

Figure 9 is a cross section of the mounting of an anemometer control unit on the handlebars Figures 10,11, 12,13 are raw data from calibration excursions under various conditions.

Figure 14 shows calibration curves from calibration excursions under various conditions Figure 15 shows a journey with groundspeed and airspeed logs.

Figure 16 shows a journey with groundspeed and wind component.

Figure 17 A shows the effect of tilt angle of an alternative design anemometer 10 device on calibration curve with a road bicycle.

Figure 17 B shows the effect of tilt angle of an alternative design anemometer device on calibration curve with a hybrid bicycle.

Figure 17 C shows the calibration curves for a road bicycle and a hybrid bicycle overlaid for an alternative design anemometer device.

Figure 18 shows a journey with groundspeed and airspeed logs for an alternative design anemometer device.

Detailed Description of the Embodiments For the purposes of explanation of the current invention, two types of bicycle will 20 be taken as exemplary. It will be clear that the same considerations will apply to numerous other bicycle types or other human-powered vehicles. However, these two types, the"hybrid"road bicycle mountain bicycle, and the road bicycle, are the two most popular configurations in current use. Thus, Figure 1 is a schematic representation of the air movement patterns that would be encountered by an airspeed measuring device. Thus 2 represents the on-coming airflow encountered by a moving hybrid bicycle, 1, while 3 represents the rotating air movement induced by the rotation of the front wheel, 4. Figure 2 further represents a possible arrangement for mounting an anemometer 5 and its controlling unit, 6, on the handlebars 7 of the bicycle 1. The arrangement is shown schematically here, but will be shown in more detail in Figures 5-9. Similarly, Figure 3 30 is a schematic representation of the air movements that would be encountered by an airspeed measuring device mounted on road bicycle, 8. The on-coming airflow is represented by 9, the rotational flow from the front wheel 12 is shown at 10, while an additional turbulence may be induced by the more forward location of the drop- handlebars, 13, of this model. Figure 4 shows the mounting of an anemometer 5 and control unit 6 on the handlebars 12 of the road bicycle 8. Figure 5 shows a more detailed representation of a commercially available anemometer 6, with LCD display 16, key pad 17 and RS232 cable connection for computer interface, 18. Figure 6 shows a more detailed representation of the anemometer device, 5, itself, as well as it's rotating vanes, 19. Figure 7a and 7b represent, respectively, top and side views of the mounting of the control unit 5 and anemometer, 6, on handlebars 21 (not to scale). Figure 8 shows 10 clarification of the mode of mounting of the anemometer on the handlebars. Figure 8 represents a section X-X from Figure 7B. The anemometer 5 is attached by bolt 23 and secured by nuts 24 and 25 to a pipe-clamp 22, which wraps around the handlebar 21, and is itself clamped in place via the bolt 26 and lock nuts 27, 28. A similar arrangement for the control unit, 6, is shown in Figure 9, corresponding to the section Y-Y of Figure 7a. A pipe clamp 32 is secured to the control unit 6 via bolt 29 and securing nuts 30 and 31, while the clamp is secured to the handlebar 21 by bolt 33 and locknuts 34 and 35.

The foregoing Figures 1-4 are illustrative of the variability of local air flow effects on a bicyclist and Figures 5-9 are illustrative of a practical way of mounting an 20 off-the-shelf anemometer to function as an airspeed measuring device on commonly used bicycle types. These are not meant to limit the scope of the invention, and it is clear that numerous other complicated air flow patterns may occur on a multiplicity of human- powered transportation devices. Also, for simplicity, the bicyclist is not shown in the Figures 1-4, although the bicyclist will also influence the air flow patterns, indeed, in different ways depending on the stance, which may be upright on the hybrid bicycle 1 of Figures 1 and 3, or extremely far forward and low down with the drop-handlebars of the road bicycle 8 of Figures 2 and 4. Furthermore, numerous windspeed or airspeed measuring devices may be adapted for use as illustrated here, not limited to the rotating vane type, but also encompassing hot wire or Pitot tube based devices. For the sake of 30 illustration, the specific device 5 shown in Figures 5-9 is especially useful since it includes in the control unit 6 data-logging capability and means for downloading the log into a computer file. This latter data logging feature is extremely useful in demonstrating the inventive principles herein, but is optionally not present in a variant where the real- time display only is desirable. The data collected by such a device, and further data derived therefrom is illustrated in graphical form in Figures 10 to 16. Figures 10-13 show the data collected during the calibration procedure of the current invention. Figure 10 shows the time course of the airspeed recorded as a function of minutes and seconds, at 2 second intervals during excursions in back and forth along the same straight path over a course of approximately 1 mile, by a bicyclist on a hybrid bicycle. The arrows are meant to bracket the time intervals during which data was further processed. Thus, the even- 10 numbered arrows 36, 38, 40 represent the outward excursions, while those bracketed by the odd-numbered arrows 37,39, 41 represent return excursions. Pauses between excursions with no motion gave demarcation points which permitted identification of the data bracketed by the arrows 36-41. Figure 11 is the plot of a similar data set obtained over the same course by a second bicyclist with a second hybrid bicycle, back and forth over approximately the same path. Thus, in a like manner, the even-numbered arrows 42, 44,46 and 48 bracket the outward excursions and the odd-numbered arrows 43,45, 47, 49 bracket the return excursions, again, data being collected at 2 second intervals and plotted as a function of minutes and seconds. Figure 12 is similar to figures 10 and 11, with a third bicyclist on a road bicycle, also, on a day with relatively gusty winds. Figure 20 13 is the same bicyclist as in Figure 12, with the same bicycle as in Figure 12, but on a calmer day. This difference in wind conditions is apparent from the appearance of the graph profiles between Figure 12 and Figure 13. Individual forth and back excursions are shown by arrows in the same way as Figures 11 and 12. Figure 15 shows regression analysis of data derived from calibration excursions such as in Figures 11-15, with the mean airspeeds thus obtained plotted against the corresponding mean groundspeeds obtained from conventional bicycle computers in the same excursions. Figure 16 depicts a typical recreational excursion ride, with parallel groundspeed data obtained from a conventional bicycle computer and the raw data obtained from an anemometer device, as a function of time in minutes and seconds. The conventional bicycle computer records 30 values at 20 second intervals, while anemometer records data at 2 second intervals. Note the consequent"spikiness"of the airspeed values, since the anemometer can detect even

very brief gusts of wind, and also responds much faster than the conventional bicycle computer to changes of speed. The conventional bicycle computer of necessity averages data obtained over several wheel rotations and has damping that will eliminate artifacts due to individual pulses from the passage of the magnet past the detector in individual wheel rotations. The excursion was forth and back, and the turning point is apparent from the discontinuity at approximately 11 minutes on the horizontal scale. Figure 17 depicts the same excursion, but here the data is plotted as distance in miles obtained from a conventional bicycle computer, and the windspeed data has been corrected to true airspeed, according to the correction factor obtained from calibration such as in Figure 14, a running average for every 10,2 seconds points calculated, so that transients that appear in the anemometer record but not in the conventional bicycle computer record, are smoothed out. Each airspeed value is then subtracted from the groundspeed value occurring at the same time, and the data expressed as relative airspeed. It can then be directly appreciated that the wind component was favorable for the outward journey and in opposition in the return journey. Further, the wind component was less in the first and final miles of the excursion since these were in a relatively more protected area.

A further feature of the present invention is the capability of measuring power output and calorie consumption. The following considerations can be used to calculate the power to maintain a given speed, following a simplified analysis from the "Composite Arts and Science"web site http://www. cas-bikes. com/pagel5. html: Where P = Power, V = Velocity, D = sum of the drag. The drag consists of the following: Where M = Rider weight plus the bicycle, Crr = Static wheel rolling resistance coefficient, Cdr= Dynamic wheel rolling resistance coefficient (s/M), Cp = Aerodynamic

drag coefficient for a bicycle, p = Air density at a given temperature and altitude (kgmlM3), A = Frontal Area (M2), V = Speed (M/s).

Hence: The foregoing conventionally neglects any effect of wind, although wind will clearly significantly affect the value of P, especially as it influences the result by the power of 3.

In the presence of wind, the formula becomes Where Vgs is the groundspeed and Vas is the airspeed determined according to the methods of the present invention. Thus, by the inclusion of this additional variable, bicycle computers are able to make an exact determination of the power being consumed at any given speed during a journey, and thus to integrate the value into an exact number of calories consumed. The constant terms in the equation can either be included as a calculation using"typical"values, or determined empirically by the bicyclist (see, for example Radfahren 2/1990, pp. 47-49 or the web site damonrinard. com/aero/measuring. htm by Rainer Pivit). Further, the work against or gained from gravity in uphill and downhill terrain, respectively, can be included in the calculation with the help of altimeter readings and an additional term M*Sg where S is the gradient of the slope as a fractional unitless value and g is the gravitational acceleration term. The coefficients in the equation (4) may be determined empirically by setting the value of P to zero and monitoring Vgs and Vas under conditions of coasting from a significantly high speed to rest while free-wheeling, or by approaching terminal velocity under conditions of free-wheeling on a gentle gradient.

When total forces, F, acting on the cyclist are take into account, equation (2) contains additional terms for forces due to acceleration, deceleration and climbing or descending:

where S is the gradient (unitless) and a is the acceleration, and may be expressed as dVgs/dt, the first derivative of the groundspeed with respect to time. On level ground, under conditions of free-wheeling and coasting to a standstill from a significant speed, the value of F is zero and the equation (5) simplifies to A plot of Vas2 against AVgs/At will yield a straight line with slope-l/2Cpp A/M and intercept on the vertical axis- (C, + Cd,). For use in obtaining real time values during an excursion, these coefficients may then be inserted into the equation (5) to determine the transient value of the force being used at any given time. This can then be converted to units of power output by the bicyclist This can then be converted to calorie units by either integrating the power over a period of time, or multiplying the mean power by the time, and multiplying by the well-known conversion factor. The methods of the current invention thus permit the direct computation of calorie consumption directly from physically measurable parameters and with no simplifying assumptions.

Current programs for analysis of bicycle computer data, such as those of Ciclo or Polar, give estimates of wind chill factor. This also would be greatly improved by the use of airspeed rather than groundspeed.

Examples Example 1. A hybrid bicycle manufactured by Trek, Model 7200 Multitrack, weight 37 pounds, equipped with a computer manufactured by the Cat Eye company, model Cordless 2, was further equipped with a Vane Thermo-Anemometer Datalogger, Model 451126 manufactured by Extech Instrument Corporation. Mounting was as shown in 7-9. A 135 pound female cyclist of moderate athletic ability performed a set of calibration excursions according to the methods of the present invention. For each run, the anemometer was set in the record mode, and data collected continuously at 2 second intervals. At the beginning of each excursion, the bicycle computer was started, and 10 excursions were performed back ad forth over a straight, approximately level, sheltered path of 0.5 miles. Enough pause was made at the beginning of each excursion so as to clearly de-mark the time point in the anemometer record. Successive back and forth pairs of excursions were performed attempting to maintain speed according to the bicycle computer of 5,8 and 12 MPH, respectively. The actual distance, average speed and time were noted for each excursion. The raw data obtained is shown in Figure 10. For each excursion, the mean airspeed was computed and then averaged pairwise for each back and forth direction. Comparison with the corresponding averaged pairs of groundspeed values from the bicycle computer permitted the correction of the raw airspeed values to corrected airspeed values, assuming any effect of wind would be cancelled out by the 20 back and forth excursions. While there may be periodic short term variations and gusts of wind that would average out, any shift in average windspeed or direction between excursions would falsify this assumption. As will be seen, the assumption appears to have been supported by the consistency of the data obtained.

Example 2. A hybrid bicycle manufactured by Cannondale, Model H300, weight 28 pounds, equipped with a Cat Eye computer, Model Enduro 2, and the same anemometer as in Example 1. The subject was a 160 pound male well-trained male. The excursions were 0.8 mile back and forth on substantially the same path, and attempts were made to maintain speeds of 6,9, 12 and 15 MPH. The raw data obtained is shown 30 in Figure 11, and was processed as in example 1.

Example 3. A road bicycle manufactured by Bianchi model Titanium with Mavic Criterium Aero wheels, weight 22 pounds, and equipped with a Cat Eye computer, Model Enduro 2, was further equipped with the same anemometer as in Example 1. The subject was a 120 pound female, experienced tri-athlete. The excursions were back and forth on substantially the same path, and attempts were made to maintain speeds of 9,12 and 18 MPH. The raw data obtained is shown in Figure 12, and was processed as in Example 1. It is apparent from Figure 12 that this was on a day with substantially stronger wind gusts.

Example 4. This was identical to the Example 3, except that it was carried out on a calmer day, and speed goals were 4,12 and 18 MPH.

Example 5. The various airspeeds and corresponding averaged groundspeeds from Examples 1-4 were subject to regression analysis and plotted in Figure 14. It is apparent from Figure 14 that (a) the two bicycles and bicyclists of Examples 1 and 2 gave data that are substantially collinear. Further, the duplicate sets of runs for the same bicycle and bicyclist under different wind conditions, of Examples 3 and 4, also gave data that were substantially collinear. Further, all of the least squares regression slopes were within a range of + 5%. This supports the conclusion that consistent calibration factors may be calculated. This may be on an individual basis, if high accuracy is required, or with an average default value if less accuracy is required. The actual calibration factors are given in Table 1: Data from Regression Standard slope Error Example 1 1. 226 0. 013 Example 2 1.256 0.006 Example 3 1.114 0.021 Example 4 1.083 0.010 It can be seen that the standard error of fit to a straight line is about 1 % for all except Example 3, which was the worst case. Even for that, a value with a standard error of fit of about 2% is extremely satisfactory for the purposes of exercise calculations. The use of such a calibration factor is analogous to the use of the wheel circumference for calibrating the mileage measurement in a conventional bicycle computer, and this can be done, at best, with about a 5% precision. The circumference measurement can also be applied by using convenient default values, or very precisely measured according to the desires of the individual bicyclist.

10 Example 6. The bicycle and bicyclist of Example 1 was equipped with the same anemometer, but with a Ciclo Sport computer model CM436M. An out and back excursion was made under a variety of conditions of hills, directions and wind exposures, to be more typical of a bicycle trip than the foregoing. Further, the bicycle computer and the anemometer were started running simultaneously and kept running over the total journey of about 5 miles. The anemometer collected readings every 2 seconds, whereas the Ciclo computer collected data every 20 seconds. The recorded data from both of the foregoing devices were downloaded into a personal computer and analyzed in the same spreadsheet. The raw data is plotted in Figure 15. It is noteworthy that the anemometer is capable of detecting gusts of wind of extremely brief endurance, which may be of little 20 significance to a bicyclist. The design with extremely light, low friction, vanes, means that the anemometer is extremely responsive to brief changes. Such"spikiness"can easily by suppressed by appropriate averaging procedures. Further, other designs of anemometer, such as the hot wire type or the Pitot tube type, are likely to have slower response times. In order to permit further evaluation, the airspeed values from the anemometer were averaged in groups of 10, thus permitting precise comparison with the groundspeed values taken at the same time. The averaged values were further corrected according to the calibration factor obtained in Example 1 and shown in line 1 of Table 1.

Finally, a relative windspeed value was obtained by subtracting each airspeed from the corresponding groundspeed value. Thus, for a windless condition, the value would be 30 zero, a tailwind would yield a negative value and a headwind would yield a positive value. The results of this analysis are plotted in Figure 16 as a function of the distance determined by the bicycle computer. The results show a tailwind in the last half of the outward journey, a headwind in the fist half of the return journey, a discontinuity at the turnaround point.

Example 7. An alternative design anemometer device became commercially available subsequent to the collection of the foregoing data in Examples 1-6. This device, the ADC Summit is manufactured by the Brunton Company of Riverton, Wyoming, USA. Breifly, the device is a compact, rugged, weatherstation with capability of reading and datalogging altitude, barometric pressure, temperature and wind speed. This 10 alternative design anemometer device was more suited to use on a bicycle than that manufactured by Extech and as used in Examples 1-6. The device could be mounted on the handlebars without the system of bolts and clamps shown in Figures 7-9. The hand- strap was removed from the Brunton device and it was simply mounted on the handlebars with cable ties threaded through the slot for the removed strap. Compared with the Extech device, the Brunton device did not require a second unit to be mounted on the handlebars. Further, the vanes of the Brunton device were 1 lmm maximum diameter, compared with 52 mm for the Extech device. The Brunton device has three"stator" blades which are approximately 3 mm deep, whereas the Extech has, in the equivalent position, three support pillars which are semicircular cross-section, about 2 mm diameter.

20 The total path for airflow in the Brunton device is 14.5 mm, compared with 30 mm for the Extech device. Thus, the air-flow path relative to the vane diameter is 1.45X for the Brunton and 0.58X for the Extech. It is possible that any or all of these radically different geometries will yield different responses to local turbulent air flow conditions. Because of its geometry, the LCD scrren of the Brunton device is not readable from the position of the cyclist when mounted vertically. The vertically mounted position would be expected to be optimal for airspeed measurement. The device was therefore calibrated at different mounting angles, with the CicloSport bicycle computer of Example 6 measuring and logging the goundspeed in parallel. The Brunton device was calibrated once mounted vertically and again at 45° tilt. The 45° tilt was optimal for reading in real time by the 30 cyclist. The calibration runs were performed exactly as in the preceding examples. The results are shown in Figures 17A and 17B for two different bcycles. The bicycle of Figure 17A was of the brand Mercier, and was a roadbicycle comparable to that of Figure 14. The Cannondale was identical to the hybrid bicycle of the same example. It is clear that there is surprisingly little effect of tilt at this angle, even though the aperture presented to the direction of flow would be reduced by a factor of about 1.4. The data of Figures 17A and 17B is re-plotted in Figure 17C to show the difference between the two bicycles. The curves almost overlay, so that, with this geometry of anemometer device, the bicycle-to-bicycle variation is far smaller than that shown in Figure 14. This example leads to the conclusions that: The mounting angle of the Brunton device is not critical, so that it can be mounted for optimal readability; and that bicycle-to-bicycle variation is 10 sufficiently small that a standardized correction could be applied. Thus, the geometric design of the anemometer can be used to compensate for the air turbulence effects, as an alternative to the algorithm described in the preceding examples.

Example 8. The Mercier bicycle of example 7 was equipped with the same anemometer and bicycle computer as in Example 7. All functions on both the Brunton device and the CicloSport device were activated during an 8 mile trip. The Brunton device was set to collected data every five seconds, while the CicloSport device collected data every 20 seconds. Concurrently, air temperature and altitude were recorded in both devices. The logged data from both were downloaded into a computer and combined into one file. The calibration data of Example 7 was used to convert the apparent airspeed file 20 into a file of true airspeed. The data for every successive seconds were averaged to create a dataset that could be compared directly with the distance and groundspeed file from the CicloSport device. Both groundspeed and true airspeed are plotted together in Figure 18. Further, the difference between them, the relative windspeed, is also plotted in the Figure 18. The sharp troughs represent brief stops at turnaround points. In each case, even though the wind was gusty, the relative wind showed a reversal in sign at the turnaround. Further data on the overall trip was obtained from the software provided with the CicloSport device, and this was combined with averaged data from the Brunton device. The two sets were pooled and an overall summary of the trip is shown in Table 2: Date 8/29/2004 Start time 5: 18 PM Duration 35: 41: 00 Time in motion 34: 00: 00 Time standing 1: 41 Distance 7.97 miles Air miles 8. 13 miles Maximum altitude 751 ft Minimum altitude 715 ft Ascent 62 ft Descent 72 ft Average groundspeed 13.53 mph Average airspeed 13.81 mph Maximum groundspeed 19.01 mph Maximum airspeed 20.19 mph

The"air miles"is a value obtained by multiplying the average true airspeed by the time spent in motion. This number is a useful means for computing net effort, since it indicates that, in spite of the fact that the wind was light and gusty, there was overall a net unfavorable effect of wind which can be seen from the air miles being larger than the distance. Conversely, if the air miles were less than the distance, there would have been a net favorable effect of the wind. In the absence of the device of the present invention, the bicyclist would not have been able to know whether he had done more work or less work in the face of a gusty and irregular wind. The same situation is reflected in the average airspeed being greater than the average speed. This example shows how the computed data from the device of the present invention usefully supplements that obtained by conventional means.

It would be obvious to one skilled in the art to create algorithms for computing this relative windspeed in real time and displaying it directly in the bicycle computer

display. It could be displayed in a variety of formats, analogue or digital, and would be of great utility for bicyclist seeking to minimize the adverse effect of a headwind. Such a precise indication is not available from any subjective feeling of wind. It can also be useful for seeking the optimal position in a path when an individual is drafting behind a group.

The preceeding examples are intended only to illustrate the utility and practicality of the current invention, but should in no way be construed as limitations to the scope.

Further, the specific configurations used therein were chosen for practical reasons of availability off the shelf and capability of data logging. It is obvious that other devices and configurations may lend themselves better to a full integration of the invention into a unitary bicycle computer system. Also, the same methods and principles are equally applicable to other means of human-powered transportation.