Description as a Tweet:

Inspectio is a navigation buddy for the visually impaired. This user friendly, low cost, hardware-software integration is an excellent aid for the blind to walk independently. #Accessibility #HelpingTheVisuallyImpaired


Our project is an accessibility product which moves around a specific disability. The mere idea of helping a community which otherwise has a lot of challenges in carrying out their day-to-day activities was our inspiration for the project.

What it does:

The project senses obstacles in the path of the user and produces sound alerts to help them navigate their way.

How we built it:

We connected the sensors to the Arduino board and programmed the set-up to calculate the distance of the obstacle from the sensor. Then we programmed the Camera for CV. All of this got connected centrally through the Raspberry Pi.

Technologies we used:

  • C/C++/C#
  • Python
  • Arduino
  • Raspberry Pi
  • AI/Machine Learning
  • Other Hardware

Challenges we ran into:

We waited for 10 hours to get a 3-D printout of the belt. Took us a considerable time ideating the design. However, the final product of the belt didn't work well in terms of flexibility along the hinges. The belt broke! We had to make last minute changes with the design. Along with this, the CV took a lot of effort to turn out the way we wanted it to.

Accomplishments we're proud of:

We could figure out how to implement Tensor Flow and the Raspian operating system. Bridging the Raspberry Pi and Arduino was also a challenge which we could finally work out.

What we've learned:

Our group had members with different levels of understanding of computer science. Some of us had no background of hardware and learnt how to make connections, solder wires etc. Others, who knew a bit about hardware but not necessarily programming, learnt the semantics of the Arduino and Raspberry Pi codes. Overall, we learnt how to 3D print, form a bridge between Arduino and Raspberry Pi, work with Tensor Flow and CV. More importantly, we learnt how to come up with backup ideas which happened when our 3D belt didn't work out. Along with this, it was a great experience to figure out how to integrate the different skills of all team members to create the project.

What's next:

We hope to add motors to the sensors to cover a wider field of detection. Our camera detection needs to be coordinated with the audio for better use of CV. Along with this, we need to make our design more compact. We need to integrate better processing chips over our current Raspberry Pi set-up. This could help us make the object detection system faster and making the design compact. Wireless connections could help set-up sensors in areas which would otherwise be difficult with wires.

Built with:

Raspberry Pi, Arduino, ultrasonic sensors, speakers, camera, wiring, programming languages- Python, C

Prizes we're going for:

  • Best Hardware Hack
  • Best STEM Hack

Team Members

Tergel Molom-Ochir
Sarthak Mital
Roshan Lamichhane
Eric Chien
Aarushi Sharma

Table Number

Table 23

View on Github