Seeing Eye dogs are important aides for many members of the visually impaired community. Yet for some individuals, having a live service animal is an extreme burden in terms of cost, care, and more. A robotic alternative could better suit the needs of some users, with less maintenance, greater longevity, and increased individualization. This research aims to explore the development of robotic service dogs (RSD) for people with visual impairments. A traditional rigid harness was adapted with vibration motors, creating a familiar feeling tool with a new type feedback. This type of vibration feedback could also be implemented on a traditional, live Seeing Eye dog with a rigid harness.
Background Research
Haptic feedback has significant usages both outside of robotics and within it [1] - [7]. The most popular applications are with smartphones, gaming applications, and virtual reality systems[3], [8]-[10]. More recently, haptic vibration has been involved in medical robotics to give surgeons feedback when conducting robotically assisted operations [6], [10] - [12]. There has also been research into using haptic feedback as a way to assist individuals with difficulty processing sensory information [7]. Thus, the understanding of haptic vibration feedback is ever growing for usages that range from medical, to recreational, to daily life. Haptic vibration feedback has significant research to aid in adapting service technologies for individuals with visual disabilities [13] - [18]. There are several methods of implementing haptic feedback. One popular method is through wearable technology such as bracelets, gloves, or belts [13], [17], [18]. The placement of the device varies with what method of sensing is used. When using something attached to the wrists or hands of the user, there is more movement to try to account for with a typical human gait than mounting close to the user's center of mass [13], [18]. Thus some researchers attempt to place sensors in places with more predictable and limited movement, such as integrating with glasses [13]. Often, these projects force the individual to learn totally new technologies and systems, rather than updating familiar technology. Finally, this research builds off of past focus group session feedback [19], [20]. This includes creating a RSD AI voice module for a Go1 quadruped and developing a "smart" cane with LiDAR detection [19], [20]. This variation allows for individuals to select their own preference rather than hyper-focusing on one type of assistive device.
First step of the project was figuring out the scope of the project. Dr. Hament and I talked extensively about the best ways to organize this project, what were good goals, what would be unachievable in this time frame, and most importantly, what success would look like. From this, I developed technical design requirements.
From this, the following sketches and CAD designs were developed.
The Unitree Go1 was used as the platform for this iteration of RSD device. A base plate was modeled in SolidWorks and 3D printed to connect to the back plate of the dog. This base plate allows for connection to the 5 step articulating hinges as well as the electronics box. The hinges allow for customization of the angle of the rigid harness. This allows the user to adjust the handle height and the horizontal distance to the quadruped. The electronics housing box contains all the necessary sensors and controls. The motors are mounted in the handle of the rigid harness. This is shown below.
The system was designed to provide distance-dependent haptic feedback using ultrasonic sensing and DC motor actuation. Three ultrasonic sensors were used to detect the proximity of objects, and each sensor was paired with a corresponding DC motor. The intensity of motor vibration was modulated based on the measured distance, such that closer objects produced stronger vibrations. This system focuses on close-range obstacle avoidance. The furthest distance to alert the user is 75 cm, and the closest is 3 cm.
Electrical and Coding Configuration
Each ultrasonic sensor was connected to the Arduino using a trigger (TRIG) and echo (ECHO) pin pain. The sensors operate by emitting an ultrasonic pulse and measuring the time required for the reflected signal to return, enabling distance calculation based on the speed of sound. The three DC motors were controlled using two motor driver modules. Each motor driver channel consisted of two digital input pins for direction control, and one enable pin for speed control via pulse-width modulation (PWM). The enable pins were connected to PWM-capable pins on the Arduino to allow continuous variation in motor speed. An external battery pack was used to power the motors. All system components shared a common ground to guarantee consistent signal referencing and reliable operation.
Distance measurements were obtained using a time-of-flight approach. For each sensor, the Arduino generated a 10µs pulse on the trigger pin, causing the sensor to emit an ultrasonic wave. The duration of the returning echo pulse was measured using the pulseIn() function. This duration was converted to distance using: d = 0.5*t*v
Where t is the measured echo time and v is the speed of sound in air, 0.0343 cm/µs. The division by two accounts for the round-trip travel of the signal. To reduce interference between sensors, measurements were taken sequentially with short delays between each reading.
Each motor was controlled independently using a unidirectional drive scheme. Direction pins were set to maintain forward rotation, while speed was controlled through PWM signals applied to the enable pins of the motor drivers. The Arduino’s analogWrite() function was used to generate PWM signals with values ranging from 0 to 255, corresponding to 0% to 100% duty cycle. A value of zero resulted in no motor activity, while higher values increased the average voltage supplied to the motor, producing stronger vibration.
A mapping function was implemented to convert measured distance into a motor speed command. Two threshold parameters were defined: a minimum detection distance and a maximum detection distance. Distances greater than the maximum threshold resulted in the motor being turned off, while distances below the minimum threshold produced maximum motor speed. For distances within this range, a normalized “closeness” value was computed and mapped to a PWM output range. To enhance perceptual sensitivity, a nonlinear scaling function was applied to the normalized distance, emphasizing stronger changes in motor output at closer distances. This approach improved the responsiveness of the system and made variations in proximity more perceptible to the user. Additionally, a minimum active motor speed was enforced to overcome the dead zone of the motors, ensuring that once activated, each motor produced a noticeable vibration.
The control logic was implemented in the Arduino IDE using a modular structure. Separate functions were defined for distance measurement, distance-to-speed conversion, and motor actuation. This modular design improved code readability and allowed for easier parameter tuning.
People
- Dr. Hament - my mentor who read my papers the day that they're due to help me get published
- Justin - who dealt with the electronics and coding aspects so that I didn't have to
- William - who helped me brainstorm testing
Software
SolidWorks - 3D modeling PrusaSlicer - prepping models for 3D printing Arduino IDE - coding Microsoft Teams + OneNote - communication and documentation
Machinery / Physical Tools
PRUSA MRK4 - 3D printer in the EGR building Clark Prototyping Lab in the Engineering Building Maker Hub - Downtown
The next step for this research is to conduct another focus group session and to receive feedback from members of the visually impaired community on this device. This focus group will allow for interaction with the prototype developed and a group Q&A session and discussion. This will allow for a better understanding of how future work can best support individuals and enhance their comfort when using such a device. Additional testing will also be conducted prior to the focus group including testing the 2D feedback field, identifying coordinates of objects in the sensor field, and scaling haptic response of the different motors to give a better sense of obstacles in 2D. This testing will elevate the quality of the prototype that will be presented to the focus group, allowing for a better understanding of what changes should be implemented. Further testing should include leading a member of the research team through an obstacle course based on the feedback detected. This research shows the significant work done in adapting a rigid harness to return haptic vibration feedback to a user with visual impairments with the ability for individual customization. Yet, there is more work to be done to ensure a safe and well tested mechanism.
Biblography
[1] H. W. Tappeiner, R. L. Klatzky, B. Unger, and R. Hollis, “Good vibrations: Asymmetric vibrations for directional haptic cues,” in World Haptics 2009 - Third Joint EuroHaptics conference and Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems, 2009, pp. 285–289. [2] J. R. Blum, I. Frissen, and J. R. Cooperstock, “Improving haptic feedback on wearable devices through accelerometer measurements,” in Proceedings of the 28th Annual ACM Symposium on User Interface Software and Technology (UIST ’15). New York, NY, USA: Association for Computing Machinery, 2015, pp. 31–36. [Online]. Available: https://doi.org/10.1145/2807442.2807474 [3] X. Tang, A. Talhan, H. Nisar, R. Lindgren, and M. Lira, “Good vibrations: Feeling and interpreting haptic feedback,” in Proceedings of the 19th International Conference of the Learning Sciences (ICLS 2025). International Society of the Learning Sciences (ISLS), 2025. [Online]. Available: https://doi.org/10.22318/icls2025.360851 [4] A. Chang and C. O’Sullivan, “Audio-haptic feedback in mobile phones,” in CHI ’05 Extended Abstracts on Human Factors in Computing Systems, ser. CHI EA ’05. New York, NY, USA: Association for Computing Machinery, 2005, p. 1264–1267. [Online]. Available: https://doi.org/10.1145/1056808.1056892 [5] W. H. Hampton and C. Hildebrand, “Haptic rewards: How mobile vibrations shape reward response and consumer choice,” Journal of Consumer Research, vol. 52, no. 5, pp. 1043–1070, 2026. [Online]. Available: https://doi.org/10.1093/jcr/ucaf025 [6] J. O’Neill, D. A. O’Neill, W. J. Lewinski, and M. E. Hartman, “Toward a taxonomy of the unintentional discharge of firearms in law enforcement,” Applied Ergonomics, vol. 59, pp. 283–292, 2017. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0003687016301661 [7] K. Krishnan, N. Jomhari, R. Kumar Ayyasamy, S. Abdul Kareem, and S. Krishnan, “Haptic feedback: An experimental evaluation of vibrations as tactile sense in autistic people,” IEEE Access, vol. 12, pp. 81 088–81 104, 2024. [8] R. Nayak, S. Mishra, P. Rane, P. Narayan, and N. Prabhu, “Development of a wearable haptic guidance system for visually impaired using gps and vibration feedback,” Applied Sciences, vol. 12, no. 9, p. 4686, 2022. [Online]. Available: https://doi.org/10.3390/app12094686 [9] C.-M. Wu, C.-W. Hsu, T.-K. Lee, and S. Smith, “A virtual reality keyboard with realistic haptic feedback in a fully immersive virtual environment,” Virtual Reality, vol. 21, no. 1, pp. 19–29, 2017. [Online]. Available: https://doi.org/10.1007/s10055-016-0296-6 [10] K. Bark, W. McMahan, A. Remington, J. Gewirtz, A. Wedmid, D. I. Lee, and K. J. Kuchenbecker, “In vivo validation of a system for haptic feedback of tool vibrations in robotic surgery,” Surgical Endoscopy, vol. 27, no. 2, pp. 656–664, February 2013. [Online]. Available: https://doi.org/10.1007/s00464-012-2452-8 [11] C. Pacchierotti, D. Prattichizzo, and K. J. Kuchenbecker, “Cutaneous feedback of fingertip deformation and vibration for palpation in robotic surgery,” IEEE Transactions on Biomedical Engineering, vol. 63, no. 2, pp. 278–287, 2016. [12] I. El Rassi and J.-M. El Rassi, “A review of haptic feedback in tele-operated robotic surgery,” Journal of Medical Engineering & Technology, vol. 44, no. 5, pp. 247–254, July 2020. [Online]. Available: https://doi.org/10.1080/03091902.2020.1772391 [13] S. Scheggi, A. Talarico, and D. Prattichizzo, “A remote guidance system for blind and visually impaired people via vibrotactile haptic feedback,” in 22nd Mediterranean Conference on Control and Automation, 2014, pp. 20–23. [14] D. Kleinberg, R. Yozevitch, I. Abekasis, Y. Israel, and E. Holdengreber, “A haptic feedback system for spatial orientation in the visually impaired: A comprehensive approach,” IEEE Sensors Letters, vol. 7, no. 9, pp. 1–4, 2023. [15] S. Khusro, B. Shah, I. Khan, and S. Rahman, “Haptic feedback to assist blind people in indoor environment using vibration patterns,” Sensors, vol. 22, no. 1, p. 361, January 2022. [Online]. Available: https://doi.org/10.3390/s22010361 [16] S. Kammoun, C. Jouffrais, T. Guerreiro, H. Nicolau, and J. Jorge, “Guiding blind people with haptic feedback,” 01 2012. [17] R. K. Katzschmann, B. Araki, and D. Rus, “Safe local navigation for visually impaired users with a time-of-flight and haptic feedback device,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 26, no. 3, pp. 583–593, 2018. [18] S. T. H. Rizvi, M. J. Asif, and H. Ashfaq, “Visual impairment aid using haptic and sound feedback,” in 2017 International Conference on Communication, Computing and Digital Systems (C-CODE), 2017, pp. 175–178. [19] A. J. Keefe and B. Hament, “Artificial intelligence (ai) voice module for robotic service dog,” 2024 Systems and Information Engineering Design Symposium (SIEDS), pp. 296–300, 2024. [Online]. Available: https://api.semanticscholar.org/CorpusID:269953964 [20] M. Dion and B. Hament, “Designing an interactive cane for the visually impaired,” 2025 Systems and Information Engineering Design Symposium (SIEDS), pp. 173–178, 2025. [Online]. Available: https://api.semanticscholar.org/CorpusID:279172297