Project 15. Perceiving the Unseen for Enhanced Tool Use.

Should this project be considered for the Best Project award? Yes.
Should this project be considered for the top 3 project awards? Sure.

On a scale of 1 to 10, how would you rate the overall organization/clarity?
8/10

On a scale of 1 to 10, how would you rate the overall project idea? 
7/10

On a scale of 1 to 10, how would you rate the overall research
contribution of the project idea, methodology and/or results?
8/10

Overall, the project report communicates the work very well. The variety of testing conditions and the outcomes are described in great detail.

The project sought to develop a sensory substitution device for communicating depth data recorded by a Kinect sensory through vibrations. The project resulted in a working system that was evaluated in a variety of test conditions.

It is nice to see a sober and honest assessment of the overall outcome of the project. Perhaps, a system intended for use a car that could only work at night was a bit too much on the awesome side.

Still, the project produced a working system that did what it had to do. Certainly, this was an impressive engineering effort. It was also nice to see a more extensive review of previous approaches. Perhaps, better familiarity with the state of haptic technology and sensory substitution systems intended for mobile use at the start of the project could help to avoid some of the issues. It was refreshing to see renewed focus on attention in the final report.


Regarding the future work, this reviewer can only to suggest to also add a working LIDAR system as an option together with the ultrasound sensors. LIDAR sensors work at night, are much more precise than the Kinect and ToF cameras in general, and are available in weather-resistant covers.
For communicating data, it might be still possible to use sound rather than vibrations. This will surely simplify the design of the system. Even though sound cannot communicate the wealth of visual information, it can be used to transmit signals analogous to the ones that used for the vibrations in the VFSS that was developed. Such a system could be simply a proximity alert, which would certainly help to avoid incidents during parallel parking and reduce the number of fatal incidents with children. Certainly, an acoustic system would be much easier to deploy than the VFSS.

Finally, for the ability to detect the "road" pixels and distinguish them from non-road, a model is described in [1], which was used for the mobile robot Stanley who won the DARPA Grand Challenge. This refrence can be a good start for research in this direction.

Keep up the good work!

[1] Thrun et al. The Robot that Won the DARPA Grand Challenge. 
Journal of Field Robotics vol. 23, pp. 661--692 (2006).