Navigational Aid for runners who are blind - TRIDENT
As a research assistant at the university, I worked on this project to develop a prototype for runners who are blind. This was done in collaboration with another university and a large automobile manufacturer as part of the Mobility for All project. TRIDENT stands for Tri-Positional Detection and Navigational Technology. The key features of the system include enhanced awareness of surroundings, detection of obstacles above and below the waist, modes for long and short range detection and choice of output. These features have been determined from interviews conducted with blind joggers and runners.
- Categories: Accessibility, Wearable Technology, User-Centered Design
- User Base: Runners who are blind
- Project Timeline: 2020 - Present
- Team Members: 4 students (Anirudh Nagraj, Tristan King, Apoorva Bendigeri, Mei Lian Vader) and 1 faculty (Dr.Ravi Kuber)
- Research Methods Used: Literature Reviews, Interviews, Affinity Mapping, Scenario Development, Competitive Analysis
As the lead user researcher, I successfully helped build the prototype to aid blind runners by conducting user research.
This low cost device consists of a variety of sensors to offer a wider detection range. The output works on haptics and can be used by Deaf and Hard of hearing people as well. The users do not have to spend money on devices that are expensive and bear only one sensor as this prototype is a handy alternative . We anticipate that people in Low-and Middle income countries would be able to afford the device should it hit the market. This device can be used for regular walking as well, extending from the runners . All users liked the idea of a non-intrusive belt that can be used to detect obstacles. We have made this belt as non-intrusive as possible. Majority of the users feel that the belt is lightweight and can be used as an alternative to the devices currently available.
From initial discussions, we observed that the main problems arise when obstacles with moving obstacles and ones at head height. Often, they find it hard to concentrate on the path because of the obstacles. To alleviate some of these standing problems, our team at the University of Maryland, Baltimore County decided to build a device for blind runners through the user centered approach . The project was done under the guidance of Dr.Ravi Kuber. The base requirements is that users want a device that relies on haptics and uses sensors to detect obstacles. In addition to that, the requirements point to the devices being non-invasive and portable so that it does not impeade their run in any way possible. Some of the questions that we tried to answer were :
We recruited participants using the snowball sampling technique.
Demographic: Individuals who are blind and go on runs. They can be first timers, regular marathon runners, or run with a tether.
Total number of users involved throughout the project: 11
Devices: Mobile (Android and IOS) and others (BuzzClip, SUNU Band, WeWalk Smart Cane, etc.,)
Study to assess the manner in which participants interact with technology while running and develop a prototype for the same using the human-centered design process. Done in phases.
The team collected several works related to prototypes for individuals who are blind. This rigorous search helped the team to understand the development from other researchers' perspectives. I populated and read through research papers to gain insight and information on the given topic, identify gaps in previous research. Overall, we wanted to have a solid background for our study. The literature study also helped provide an overview of some of the specifics used in assistive devices (Sensor information etc.,). Seeing that many researchers built prototypes as walking assistance and not running was a milestone for our project.
- Research related to running is at a premium. Researhers have focussed on aiding individuals who are blind while walking from one place to another.
- Research indicates the presence of more than one kind of sensors located at different regions to tackle dynamic obstacles.
- An array of sensors is used as a feedback mechanism post obstacle detection.
- Research relates to augmenting the experiences of the white cane or building a device in conjunction with the cane.
The team looked at some of the different products available on the market. We even ordered a couple of devices (The Buzzclip, SUNU Band) to try out and give a different opinion. The competitive analysis gave us the strengths and weaknesses of the various products used by people who suffer from vision impairments. Trying out the devices gave an understanding of the features and functions. Trying out the products also helped inform the design process. User reviews helped us understand what the users think about the different products.
The BuzzClip is a portable device with a single sensor, but it is expensive. The BuzzClip may fall off the clip-on object and cause problems to the person using it.Other devices like the SUNU BAND have a high learnability curve and in addition to the application that comes with it. The minimum pricing is in the higher $200 range. Not many users go out of their comfort zone of using the white cane and buy specific devices.Other devices were not tried but searched online. That can be considered a drawback in the competitive analysis because the researchers do not have the first-hand experience trying out the device.
Multiple scenarios were developed over a period to ensure that major situations are covered and the prototype addresses them. We believe that by creating the scenarios, we could generate empathy for our user base (runners who are visually impaired) and design, and develop the most optimal solution. Scenarios also made us understand the key features of our device, where it may be used and also shed some light on the environmental condition that may hamper it's usage. We created scenarios because we believed that it would help us in understanding environmental specific motivational factors and guide user thought process.
- Scenarios guided us in the design process in the sense that it allowed us to empathize with the users related to specific situations
- Shed some light on the probable limitations of the device.
- Enabled us to identify hard-to notice user needs.
- Scenarios gave us the challenges faced by the users and how we could tackle the same.
- Scenarios also indicated the motivations and thought processes of the users under consideration
We created the affinity diagram exercise to populate the themes, arrange thoughts from brainstorming into clusters and note what was important to the development of the prototype. Once we had all the data points ready, the team sat down and created an affinity diagram to draw themes and prioritize them. Collaboration was done on an online tool called MIRO . Affinity Diagram also helped us organize the ideas about the possible solution for the users and allowed the team to observe the ideas in our study about the prototype to be built. It gave an insight on the way each researcher in the tem thought about the device, solution, and the environment as a whole. Although time consuming, we gained valuable insights on what to prioritize going forward. This method was chosen because several people combined their research before the development of the prototype. We did the affinity mapping exercise to populate the themes and note what was important to the development of the prototype.
- Multiple obstacle detection: Participants want an array of obstacles to be detected and not just a single obstacle.
- A non-intrusive device that is multimodal: Participants do not want a device that impedes them in any way either physically or mentally in any way while running
- Presentation of multiple haptic feedback: Participants are looking at receiving an array of feedback in the form of haptics while approaching an obstacle.
- Detection of obstacles: Participants look forward to learning about the obstacle from a short distance away.
- Weight Constraints: Participants do not want to carry something heavy while running as it impedes their run.
- Terrain: Participants want the device to work in multiple terrains, and not just track and field.
Prototype Design and Development
We were now at a stage where we could build the prototype based on the insights. We chose 3 different sensors (Infrared, Ultrasonic, and LIDAR) to incorporate into the device, and from the research material gathered, we decided to go with a RUNNING BELT. As one of the two prototype developers on the team, I set myself a deadline to get everything up and running. Within two months, it was ready for review.
The sensors are integrated in a way to mirror what we were working towards – A running belt worn by runners who are blind to detect obstacles. Velcro is used for ease of use. A headband is an add on to the device to detect obstacles at head height and has it's output generated on a wrist band to avoid head related issues.We named the device, the TRIDENT. TRIDENT stands for Tri-Positional Detection and Navigational Technology.
Users Interviews to check for Feasibility
As a team, we want to develop a suitable device to fit our target audience. So we interviewed participants ranging from first-time runners to seasoned marathon runners about the prospective device as a concept interview based on the research. The feedback was collected, analyzed, coded for reference. We conducted the context interviews to inform the device's feature decisions based on the users' feedback.
It depends on how well you know the trail. I’ve gotten to the point where they just. Give me a really short warning. I’ll be OK because I’m prepping for it anyways. If that makes sense.
The DeafBlind cannot hear the bells as a location indicator. Again the SUNU band already does this and it was tested in a crowded race type environment
- The participants mentioned that if the weight is negligible while running, it would be ideal for them to dorn the device.
- A single sensor would not cut it in environments where wide sensing would be required.
- Participants have demonstrated that the device may not be used entirely for running. For instance, one participant voiced concern about using any device in a crowded marathon, while others mentioned the difficulty with information bombardment.
- Participants mentioned that the device could come on its own and be useful in obstacle detection given the array of sensors and the varied sensing range.
If we can move on to kind of more specifics of how the device responds. So, if it's clear that we have awareness of the user you can get objects in front of you to the sides and behind you.
Yeah, I guess my question is, how we find the vibration. Is it just going to drop off if you move to the left or right of the obstacle?
How heavy the pieces are and how bulky it is but I think that's something to definitely consider. Especially if it's a battery-operated.
Obstacle course evaluation
Obstacle course evaluation
We will be running an evaluation study with an obstacle course to draw further insights for future iterations. You can find the evaluation document here .
Multivariate Testing to be done on the different editions of the prototype.