Automated Snow Removal

I have recently acquired a Roomba. It is a robot that can automatically vacuum the floor. I have to say it works fairly well. But it got me thinking, what other tasks could similar technology be applied to.

There already exists similar products for the purpose of washing floors, cleaning pools, and mowing lawns. However, the coming winter made me think, what if there was a robot that could remove the snow from your driveway for you.

Snowblowers greatly ease the task of removing snow from the driveway, but they still require you to work in the cold elements.

Creating something that could accomplish this task would require overcoming a number of issues. Because of how daunting these problems may seem at first, I am going to take some time to address the major problems as I see them one by one.

The first problem is that outside there are no walls to provide guidance for the device. A method of keeping the snow removal on track would be needed.

The best way to accomplish this, as far as I am concerned, is by installing a guidance system along side your driveway. Imagine a system similar to that of an invisible dog fence. Systems like this are already used to guide snow plows along dangerous mountain roads where visibility can not be relied upon.

As far as how the snow removal robot would actually remove the snow, it could function just like any actual snow blower. However, if this method was used, it introduces a new problem. Without a person pushing it, the robot would need a powerful drive system. Instead a lower power scaled down system would likely be better. Rather than trying to get the snow all in one pass, it would be easier to do it in multiple passes. This would not require as strong as a drive system.

However, once again that introduces a different problem, the robot would have to be light enough not to crush the snow into ice during an early pass.

With this in mind, it seems that the robot would need to be designed to remove the snow as it is falling before it has had a chance to pile up. This would allow it to be have wider weight limitations, have a lower power drive train, and have a lower powered snow removal system.

The last final problem that I foresee is safety. You would not want a snowblower to be driving around by itself with no concept of where any nearby humans were located. I also feel that this is the largest problem. Even if a product could be made that solved the other issues, it would be meaningless if it was horribly unsafe.

Any method of visual detection would be unreliable in adverse weather conditions (such as when it is snowing) and thus must be discounted. The system can also not rely on any sort of transmission device attached to people (because people may not be wearing them).

A workable answer would be a bumper system (similar to that of the Roomba). A bumper would be in front of snowblower robot and would detect collisions. The system would then be programed to stop the snowblowers actuator upon collision with an object.

The bumper would need to be a foot off the ground to avoid being triggered by snow on the ground. It would also need to be much wider than the snowblower itself. If it was not, it could be possible for someones foot to get under the robot without their leg triggering the safety bumper.

This would not be a perfect system by any means. My proposed safety system would not protect a person lying down in front of the robot. However, snowblowers as they currently exist are not perfectly safe either.

As much as I would appreciate such a product, I am also realistic. Each of my proposed solutions include any number of engineering challenges. Even if these challenges were solved, it could be highly probable that it would not be fiscally profitable to produce and sell.