Proceedings of the Multi-Disciplinary Senior Design Conference Page 7 s3

Proceedings of the Multi-Disciplinary Senior Design Conference Page 7

Project Number: P14043

Proceedings of the Multi-Disciplinary Senior Design Conference Page 7

Smart Cane

Lauren Bell
Mechanical Engineering / Jessica Davila
Industrial and Systems Engineering
Jake Luckman
Mechanical Engineering / William McIntyre
Electrical Engineering / Aaron Vogel
Mechanical Engineering

Proceedings of the Multi-Disciplinary Senior Design Conference Page 7

Abstract

A considerable size of the population is impaired with the challenging task of navigation due to some level of vision and hearing loss. The purpose of this project was to develop a cane handle that would combine the benefits of a traditional white cane and a service animal. The solution developed was a handle that would attach to a white cane and provides directional feedback to the user using a roller assembly. The roller assembly uses four bearings that rotate in one direction or the other to indicate which direction the user should move to avoid obstacles. Ultra-sonic sensors are used to convey the information of the environment to the handle itself. The finished cane physically resembles a conventional cane therefore allowing the user to still be able to use the cane to sweep, tap, and feel the ground.

introduction

Navigating the physical world is challenging for people with any level of vision loss, it is understandably even more challenging for the estimated 42,000 to 700,000 individuals [1] who have both some level of vision and hearing loss. Current devices that are available to individuals are standard white canes and service animals. White canes provide only a slow navigation as the user must guide themselves by feeling obstacles in their path. Service animals as well come with some disadvantages, such as their cost and continuous care that must be provided. A number of first generation prototypes have been developed that detect obstacles and provide an audio or haptic warning to the user. These devices however tend to be bulky, heavy, and fail to indicate a clear direction to guide the user. Canes with audio feedback are moreover ineffective in noisy environments for deaf/hard of hearing users. A patented prototype design by Gary Behm [2] encompasses a vibration feedback that is delivered to the user based on feedback from a set of sensors. This work was the frame of reference for the project.

Design Process

The design objective of this project is a dramatically improved handle for a guide cane that provides haptic feedback that clearly informs the user of obstacles. This feedback would need to quickly respond to signals from a detection system. The cane handle would need to be designed to visually and physically resemble a conventional handle and intuitive enough to be used with minimal training. The cane would be in constant use throughout a day, therefore the comfortability of design is a significant factor. The grip diameter and weight of the handle was seen as a part of this specification. The handle would need to be powered and functioning for any tasks that the user may need the cane for during the day. Assuming that the prototype continues to the manufacturing stage, the cost of this handle for purchased parts would need to be reasonable to stay competitive with standard white canes.

The needs of the design allowed for multiple design concepts to be proposed. After much evaluation of the different designs, a “scroll” navigation handle was selected. The selected “scroll” handle concept would include a roller sub-system that contains ball bearings that will roll beneath the palm, allowing the user to feel the direction of rotation. The design entails that if the rollers rotate to the right, this would indicate that there is an obstacle to the left and the user should travel to the right and vice versa. The design intent includes a stop feedback where the roller system will rotate back and forth beneath the palm indicating the there is an obstacle directly in his/her path. Batteries would power the system and a power switch will be easily accessible to the user. The entire system would be managed by a microcontroller and printed circuit board.

From these customer requirements and design concept of the handle, engineering requirements were created, that can be found below in Table 1.

Rqmt. # / Importance / Engr. Requirement (Metric) / Unit of Measure / Marginal Value / Ideal Value
S1 / 3 / Operates at S2, S3, S5, S6 and S14 under grip pressure / psi / 3 / 5
S2 / 3 / Minimum Rotation Speed at 3 psi Grip / rpm / 20 / 20
S3 / 3 / Maximum Rotation Speed at 0 psi Grip / rpm / 60 / 60
S4 / 3 / Maximum height of bump from handle surface at 0 psi Grip / in / 0.18 / <0.18
S5 / 3 / Minimum height of bump from handle surface at 3 psi Grip / in / 0.12 / >0.12
S6 / 3 / Time from motor input signal to when the roller reaches the minimum rotation speed at 3 psi Grip / ms / 500 / <400
S7 / 3 / Circuit voltage / V / 12 / <12
S8 / 3 / Handle contents fit within handle mock up envelope / Binary / Pass / Pass
S9 / 3 / Maximum handle grip diameter / in / 1.5 / 0.78
S10 / 2 / Estimated Manufactured Handle cost / $$ / 100 / 30
S11 / 2 / Prototype cost / $$ / 750 / <750
S12 / 2 / Maximum weight of the handle / Lb. / 1 / <1
S13 / 2 / Battery life / Hours / 4 / >4

Table 1: Engineering Requirements

The average pressure a human hand exerts when gripping an object is 3 psi [3] therefore the design would need to be able to withstand this grip and still function as seen in Requirement S1 seen above. The effective height of the ball bearings from the surface of the cane handle was determined through user testing which is described later on in the paper.

ERgonomic evaluation

User satisfaction of the final prototype is an important aspect of the project. To ensure that the rollers effectively provided direction and were comfortable, a series of user tests were executed. Three tests were conducted to determine the effective roller speed, roller geometry, and height of roller from cane surface. Each test consisted of 15 test subjects testing 5 different data points. The test subjects were asked to rate the comfort and sensitivity of each scenario on a scale from 1 to 5. An analysis of variance was executed on all of the data sets collected to establish what the optimal results were. The geometry of the rollers that was found to be the most comfortable while also effectively relaying which direction the rollers were rolling in was the cylindrical shape opposed to spherical. The height of the rollers from cane surface that was found to be most effective was between 0.12 – 0.18 inches from the surface of the cane. A range of 20-60 RPM for the speed of the rollers proved to be effective in providing direction as well as being comfortable. These results found through testing were set as specifications of the design as seen above in Table 1. It was also found that 4 rollers, 90 degrees apart was effective in conveying directional feedback.

The final prototype of the handle is covered by a synthetic nylon fabric and then a tennis racquet handle cover. The nylon fabric was chosen to ensure that no outside particles interfered with the system. The tennis handle cover is moisture wicking and has a high coefficient of friction providing a good grip for the user.

Mechanical evaluation and analysis

Motor selection was a crucial process due to the fact that the motor needed to meet torque, size and rotation requirements. Theoretical analysis was done to the selected motor to estimate the torque-speed and torque-current curves. The analysis showed that the motor would be able to achieve the rotational speed requirements under the grip load. The torque-speed curve was estimated as well as the torque of the motor at known speeds. The torque that the servo needs to meet the design specifications is at least 50 oz.-in and the selected motor puts out 71 oz.-in. The motor was tested to determine its speed-power curve to prove that it will meet our design battery life requirements. Using a pulse width modulation generator to drive the motor, the current and power drawn by the motor at various speeds were recorded. Rotational speed was changed by adjusting a load that was applied to the motor. The rotational speed of the motor was also proved to meet the requirement of starting from no rotation to the minimum rotation speed under the specified grip pressure in 0.500 seconds. The small size and low weight of the motor was a key feature in providing a compact and low weight handle. In addition, the mounting platform on the gearbox provides a great degree of rigidity in integration within the handle.

To ensure that heat dissipation would not be a concern with the design, a heat analysis test was completed. The motor was loaded with 57 oz.-in, insulated and run at a speed of 20 RPM for 30 minutes and the temperature was recorded. The heat generated by the motor during the test was 1°F therefore heat was not seen as an issue for the final prototype.

The critical areas of the handle were evaluated under a worst case scenario static loading of 5 psi to ensure that the height of the rollers met the requirement found through user testing under a human hand grip pressure. Theoretical analysis exhibited that the roller assembly shafts and pins deflection is negligible in regards to meeting our bump height specification. The shaft and pins deflected less than a ten thousandth of an inch which does not hinder meeting the engineering requirements, S4 and S5, seen above in Table 1.The mechanical components should resist fatigue for infinite life. The possibility that the motor support plate would buckle under pressure was diminished after analysis was done at an excessive grip pressure. Analysis on the handle assembly deflection proved that the deflection on the assembly is very minimal and is rigid as was expected.

Special design considerations were made to ensure the handle would survive a fall (approximately 3 feet from the ground) after being dropped. To protect the roller assembly protruding from the handle surface, the ABS plastic outer collars, end cap and adapter were designed with outer diameters greater than the bump height. These ABS components will absorb all of the impact energy to prevent damaging the roller. After discussing with a subject matter expert on impacts, no further analysis was necessary since it is a very low-energy impact on very reliably tough ABS material.

The other critical design component for a drop impact was the batteries. According to the manufacturer, the batteries are able to withstand the impact of a 20lb mass dropped onto the batteries from 24 inches above the batteries, which far exceeds the impact energy the handle is expected to absorb during a 3ft drop.

A weight estimate was monitored during the design process. The estimate weight at the end of the design process was 0.59 pounds, which meets the requirement of 1 pound along with leaving some room for variance. A similar tracking system was used for the manufactured product cost establishing that the product would meet the $100 limit.

electrical design

Two lithium ion battery cells were chosen to supply power to the design. The selected batteries provide the necessary power for all circuit components including the microcontroller, power management system, proximity sensors, and motor. The diameter of the batteries was perfect for fitting within the inner diameter of the handle shell and the battery life would meet the 4-hour lifetime requirement. The batteries are rechargeable and the end cap of the handle includes an inline dc jack that will allow for easy accessible charging.

After a 5V DC motor was selected, a power source need to be created for proper operation. Using Texas Instruments TPS54239, the input battery voltage was regulated to 5V. This switching node buck converter is able to proper supply a changing load. The motor will draw more power based upon the load applied to the roller assembly, thus the buck converter is necessary for proper operation

To control the motor, Texas Instruments DRV8839 brushless H-Bridge motor controller was selected due to its size, low operating voltage and high power output. Using three 3.3V max outputs from the microcontroller, it is able to interface to the H-Bridge to control motor at 5V. Two of these H-bridge inputs are 3.3V digital inputs, giving the motor direction and brake functions. The last H-Bridge input is used to control motor speed which is accomplished using a pulse width modulation signal. Although not necessary variable motor speed was designed for future projects ensuring the magnitude as well as direction could be relayed though the haptic feedback. Figure 1 below illustrates the functionality and implementation of the H-Bridge.

Figure 1: The H-Bridge Schematic and associated truth provided by Texas Instruments

.Another buck converter was used in this design to create a 3V regulated source to power a ultrasonic sensors and provide an enable logic voltage to the H bridge. Texas Instruments’ TPS560200 was selected, due to its size, small amount of additional required componets and power consumption. This 3V supply is also able to supply any future low power devices the customer may want to include on the cane since 3VDC is a very popular operating voltage.

It was also purposed to check the voltage level of the batteries to do a rough estimation of battery life upon start up. This was done by reading the battery voltage though a unity gain buffered voltage divider. Although crude, this design was chosen due to spatial considerations on the Printed Circuit Board (PCB), time and non-essential functionality.