There has never been a sever need in the automobile industry for a car to have serious processing power until the dawn of self driving cars. Manufacturers are now faced with the unique problem of optimizing the processing capabilities within their cars. One recent innovation is trying to bridge the gap from today and tomorrow by using FPGA inside a car to monitor for system failures and control the system. The design proposes a GPU handle primary automation functions and use the FPGA to kick in when a failure is detected. The FPGA greatly enhances the total design of the system by be very reliable, low power consumption, and very low latency with decent processing power to back it up. When failure is detected, the FPGA will be able to kick in one of its 'special drivers' to handle a number of situation and control the car for 10 seconds until the driver is alerted or bring the car to a safe stop. Theses, 'special drivers' are able to handle several computer vision algorithms suc as watching the rode and controlling the car and watching the driver to see if their focus has returned. While this is not true self driving, having these fallbacks that are as reliable as a FPGA that can run constantly under almost any condition greatly enhances the factor of safety on these autonomous driving cars and will help usher in a true era of self driving cars.

Laundry is a part of everyone's life. However, not everyone's first thought of a smart appliance is a laundry machine. A smart laundry machine has many things to offer such as saving energy and more reasonable washing. Most smart laundry machines have a tough time learning the user's cloth types and turbidity changes and take a long time to train. However, using FPGA in a new designs allows for faster training with the use of genetic algorithms. Genetic algorithms is a form of machine learning which is very similar to how our own genes are created. Using mutation and crossover, the weights of the system are taken and simulated to the environment. Over about 100 different tested weights, the best 10 are taken and slightly manipulated to create another 100. This process is repeated hundreds of times until these weights are optimized. The weights are then funneled into a neural network that controls features such as the time, temperature, inflow of water, and the motor. Combine all these features on a reliable and always running FPGA and you are left with an extremely efficient washer that will wash clothes like their your clothes, not just the "standard wash" setting. This is all made possible with the large increase of training speed due to the FPGA board running all the algorithms.

One of the most challenging problems facing the computer vision and artificial intelligence community is the issue with image quality. There are frequently times when information needs to be extracted from an image but the image is distorted. With no reference picture to go from, this becomes a difficult problem referred to as No Reference Image Quality Analysis. With speed always being at the forefront of the programming world, when it comes to complex image processing, a standard CPU does not cut it. However, with the parallel processing of a FPGA, the problem becomes much easier. Using deep convolutional neural networks, a team form Beijing, China has developed a method o extract natural scene statistical features using FPGA. This problem only becomes more feasible due to the FPGA unique structure. This can be implemented in many computer vision applications across many different fields of technology.