Aqua Phoenix
     >>  Research >>  Sentinel  


5.5 Future Considerations

This technology could be greatly expanded, although the complexity would increase. One big limitation is that the camera must look at the floor in some angular fashion. This would not solve the problem of autonomous driving in the real world, where there are traffic lights, railroad crossings, bridge crossings, and stop signs. One option is to have the camera face straight ahead. With the limited visual range horizontally and vertically, however, this would not be such a good idea as well, since there is no way of perceiving cars and pedestrians to the left and right of the camera view.

Therefore a visual range increase would greatly improve the range of decisions that can be made as well as improve the accuracy of the decisions. We suggest a forward view of 90 degrees from top to bottom and a 120 degree view from left to right. The main focus would be the middle 45 degrees, as that would be the position of the road. Future algorithms could use possible shape, color, perspective, and temporal movement detection to push forth a more usable product.

The technology implemented here does have its merits in real world application. For example, waiters no longer need to carry out trays of food to customers. With lines and forks on the ground, carts could drive themselves to the correct tables and deliver the food themselves. Grand superstores like Home Depot with an elaborate setup could have carts that drive customers to the sections of the store of interest. There are several possibilities for this type of technology.

Now if some human drivers could learn from an algorithm like this