As a very "hands-on" Robotics Engineer, I've worked on many highly complex state-of-the-art robotics & computer vision systems around the world, from self-balancing robots that can climb stairs, to humanoid robots that can talk & recognize faces, to writing some of the fastest graphics and image processing software in the world. But after working on a robot for the US military, where I eventually realised the robot I was having fun building was actually intended to potentially kill women & children one day, it really changed my view on what projects I want to work on. I felt like one of those scientists who first designed the atomic bomb just for the sake of science, but as they say, with great power comes great responsibility. So in recent years I've been looking for ways to eventually put my skills & experience towards helping nature or people in need.

Luckily many of the technologies useful to robots are also useful to humans with disabilities. My main interest in robotics has always been in "robotic vision", ie: giving sight to a mobile robot. Similar techniques can also be used to give sight to a blind person! So I want to make a wearable device for blind people that helps them detect the distance to nearby objects. Many blind people use a walking stick to sense distances while walking, but a walking stick isn't too practical for sensing shorter distances such as sensing the items in front of you. So my device is not intended to replace a walking stick, it's intended to complement a walking stick and other tools, such as to help a blind person find objects on a table or cabinets in a kitchen, or find the toilet seat without having to touching it.

Being a Robotics Engineer, my initial designs were similar to how various well-funded university labs had tackled this problem: highly complex systems that need powerful computers to process 3D data from stereo cameras worn on the head, resulting in devices that cost around $10,000 just in parts.

But after living for a while in the 3rd world, and finding out that according to the International Agency for the Prevention of Blindness, 90% of the world’s 39 million blind people live in developing countries. So my main goal was to make the device useful but still affordable in the 3rd world. This required a radical change in how I tried to design the product. Instead of starting with a highly complex product and hoping it can eventually become affordable, I decided to tackle the problem from the other way: build the simplest, cheapest blind assistance device I could invent that would still be useful, and then worry about adding extra features later!

With my roughly 20 years experience in robotics, I knew there were far simpler solutions for measuring distance to objects than using 3D camera systems, such as Ultrasound, Infrared, Radar, or LIDAR. In fact all of these sensors are commonly used on robots, so I already knew the strengths and weaknesses of them all. Radar and LIDAR are highly reliable, but usually very large and expensive. Ultrasound is cheap and fairly small, but is extremely unreliable. Every time I tried to use Ultrasonic sonars for robots or other projects, they have always proven to be quite useless, especially on robots that are moving around in the real world. Quite often, even the expensive types of Ultrasonic sensors aren't reliable enough to detect someone's feet in front of a robot at times. So that leaves Infrared as the sensor of my choice! I've seen several other projects on Hackaday and elsewhere that try to measure distance to objects for the blind, but they have all been based on Ultrasonic sensors, such as Pathfinder, but as mentioned above, I'll be making one that sense distances based on Infrared.

Once the device knows the distance to the object, it can inform the user by vibrating their finger. For the initial device, it would simply vibrate more as the object is closer, but some people might prefer that the vibrations stay the same strength but change in frequency as the object gets closer.