Chapter 4. A Drone for Golfers

As we've stated a few times before, when choosing an industry to make a drone for, you have to balance the size of the market (demand), with how much money they have to spend (cost), and how easy it will be for you to fill that demand (supply).

One of the industries that is extremely popular is golf. Every year, consumers literally spend billions of dollars on golf equipment and accessories. One thing golfers go nuts about is the new technology to help their game. Sounds pretty much perfect, right?

So, where can a drone fit in? A multicopter to fly down the fairway and check the layout? Maybe, but if the golfer has played that course before, why would they need it? They wouldn't. Ok, that's a bad idea. What about a submarine drone to look for a golf ball that went into the water hazard? Not a terrible idea, but there are probably so many balls sitting at the bottom of a water hazard that finding your own ball may prove impossibly difficult. And who wants to keep...

The design

After some initial research, we found that there is already such a device out on the market. It's called the Stewart X9 Follow. However, at a list price, this item is just shy of $3,000 US!

It's certainly a high-quality product—a very sleek and beautiful design. There is no doubt that the high-end golfer would love to indulge in this item. But for your average golfer, this would absolutely break the bank.

This item also uses its own remote. The rover follows the remote, or can be piloted manually with it. We're going to use a different approach. Instead of a proprietary remote, we'll make ours follow a smart phone (via Bluetooth or Wi-Fi). This means it can be paired easily. Also, we'll include a standardized RC remote (in the event that the user's phone runs out of battery because they forgot to charge it).

Using standardized parts should allow us to bring the final cost of our unit down to around $1,000 US. Our unit may not end up as luxurious as the Stewart, but it will hit a...

Using tank locomotion

Tank style locomotion is where there are no steerable wheels. Rather, wheels on one side of a vehicle move forward, while the wheels on the other side either stay stationary, or reverse (as illustrated in the following diagram). This is also called skid-steering:

This kind of locomotion is great because vehicles can literally turn on a dime (in place). However, the drawback is that it can be jerky when turning while going forward. But as there's no sensitive equipment (just golf clubs), we don't really care if there's jostling of the clubs during turns. All we really want is to make this as inexpensive and simple as possible. You may notice that the front wheel is fixed (doesn't pivot as a caster would). This is of no concern to us as when steering occurs, the front wheel doesn't have a lot of weight on it (the center of gravity is right on top of the main wheels' axle). Therefore, the friction is minimal. If it turns out to be a problem, we can always replace the front...

Using Bluetooth to control Pixhawk with a phone

Bluetooth with Pixhawk is a snap. But just like every other bit of Pixhawk functionality, an add-on module needs to be purchased. We picked up the one shown as follows:

Of course, we had to alter the cable (just as we did before) because the cable included with this module is for the old-style plugs. We connected our Bluetooth module to the Telem 1 port as shown in the following figure:

Now, it's just a matter of binding our Android device's Bluetooth to the Pixhawk's Bluetooth. Just power on your Pixhawk and bind using the same method you would for any Bluetooth device. The software to control Pixhawk from a mobile device is called DroidPlanner 2. It's free!

We're not quite there yet, but if you're curious on how to get a Pixhawk vehicle to follow you using DroidPlanner 2, it's easy! Just click on the Follow button, and set the radius you'd like the rover to stay within. Of course, you can ignore the altitude attribute (as this is a ground vehicle...

Using LIDAR to avoid obstacles

Would you like to go for a walk on a golf course blindfolded? No? Me neither, and it's probably not good to just set a rover to run around blind either. But vision and object recognition takes a lot more computing power than we're budgeted for on this machine. So, what's the next best thing to vision? SONAR or LIDAR. Unfortunately, SONAR (just like a bat finds insects using sound waves to calculate the distance to obstacles) is considered passé for Pixhawk, and thus the implementation is quite wonky. There are I2C sonar devices, but they don't work very well (if at all) with Pixhawk 2.1 and the latest Ardupilot firmware. So, we're going to use LIDAR instead. LIDAR rangefinders use light waves (instead of sound) to calculate the distance to objects.

LIDAR 101

LIDAR simply uses a light wave to send out, bounce off objects, and then looks for it to return. Since light travels at 299,792,458 m/s, all the LIDAR module has to do is measure the time it takes to return...

Summary

In this chapter, we learned how to implement skid-steering, using a rangefinder for obstacle avoidance, and how to make a rover follow us using Bluetooth and an Android phone app. As you can see (on some) applications with Pixhawk can be far easier than they seem at first. You can pilot your rover with your phone and even use a map to lay out a course (using waypoints) for the rover to automatically drive itself. But, most importantly, your new rover will follow your phone as if it were on a leash.

Now, we'll move on to some more advanced applications for Pixhawk aircraft. Brace yourself, this is where designing, fabricating, and implementing Pixhawk gets a lot more complex. But it's also where Pixhawk really shines, and is a whole lot of fun!