I would like to know if Lidar would be a good sensor to track a baseball bat across a narrow path to display the angle of the path - see drawing.
The bat would be moving up to 100mph across the sensor. I would like to capture the distance of the bat above or below an ideal swing plane.
If not lidar what do you recommend?
It seems like it would be difficult to detect such an object (see speed below) using a single beam/point LIDAR. Do note that most of the LIDAR products we sell are geared towards detecting obstacles (i.e.: large, mostly not moving or slow moving objects) and mapping surroundings.
To track objects that move, you’d want an array solution such as some of the or BEN and TER products. These can track the distances to a 2D array of points, which allows to read the movement of objects in that array over time (difference between frames). Again, speed is an issue (see below)
This means your baseball bat would be moving at ~1760 inches/s or 4407.4 cm/s. Considering the size of a baseball bat, you’d want to be able to take measurements fast enough so that you do not miss the motion. The worst would be if only the tip of the baseball bat is visible (small target) and it is moving at ~100 mph. Assuming the tip as a circle with a diameter of at least 2.375 inches or (~6.0325 cm) - smallest I could find online quickly, we’d want to target half of that at full speed: i.e.: 1.1875 inches moving at 1760 inches/s. 1760/1.1875 => ~1482 Hz if you want to notice motions of at least half the baseball bat’s tip’s diameter.
This requires a combination of a fast enough 2D array LIDAR with enough resolution and range to work with your setup.
Some of our best 2D array LIDAR, such as the RB-Ben-06 and the RB-Ter-28 have rates of measurement of 30 and 130 Hz respectively. This is at least a full order of magnitude too slow for your case above, which means that in many cases these sensors would miss the critical moment where the bat is moving. At 130 Hz, you’d get “snapshots” of the bat’s motion (at maximum speed) of about ~13.53 inches interval, which are much larger than the bat. Therefore you’d most likely not capture the exact position you want to measure.
Faster LIDARs do exist, but the cost is prohibitive for DIYers/hobbyists and therefore we do not currently offer those currently (though that may change in the future).
Most likely a CV (computer vision) system using a fast camera and an embedded computer (such as a *Single Board Computer, like the Raspberry Pi) or a full computer. Most of the cameras we sell are 30-120 fps (or Hz) and therefore may not provide the results you want either. Just like high speed LIDAR, high speed camera and other vision systems are expensive, too… 
All being said, if having a trace of points or “snapshots” @ every 13.5 inches is enough for you to reconstruct the path of the bat after the fact then maybe that could work for you.
I suggest you look into small radar modules. Since they are being used in driver assistance systems, and autonomous cars they are becoming more common, better integrated, smaller, and cheaper. They tend to have most if not all the capabilities you need. Digikey lists several evaluation modules that you should look at: https://www.digikey.com/products/en/rf-if-and-rfid/rf-evaluation-and-development-kits-boards/859?k=radar
Distance to and speed of the bat should be fairly simple, requiring only a single “channel.” But if you want position (up, down, left, right) of the bat within field of view that will require the phased array radar, much as the arrayed lidar mentioned by scharette. Those will likely cost a great deal more, perhaps ten times as much.
Hope this helps.
That is actually a great point! We do offer some currently and you can find them here. This manufacturer has also been continuously improving the software and firmware (and adding features through those updates), so it is certainly a product to consider.
But, as @oldguy mentions, getting distance to / speed is the easy part. Getting your angle (which means a proper position in 3D or at least on a 2D plane) requires more advanced equipment. Considering the speed of the object, this equipment most likely needs very fast measurement cycles (for the entire array).
I really appreciate the time and effort you guys put in to respond to my question. I was not expecting such robust explanations. I am learning about LIDAR and now have a better understanding of how they can be used.
As you mention the speed is easy. I had already settled on two IR beams separated about 12 inches apart in front and back of the ball. The bat crossing the beams would trigger a microsecond timer that would be fairly accurate to a few millimeters. I am used to Cypress Semiconductor PSOC CPUs that can run at up to 150Mhz and reliably tick at the microsecond level.
If my math is correct, that means 4.4mm per clock tick at 4407.4 cm/s ± a few microseconds for setup and interrupt handling which is a fixed value.
Also, as you mention the challenge is the angle of the bat path. The short range radar is worth exploring. I will probably order one just to play around with this new technology. They certainly seem fast enough to cover most of the path of the bat with great accuracy if I can use the phased array to sample over time.
It really helps to talk out loud and listen to your suggestions. After sleeping on the problem, It occurred to me that all that matters is the angle at the point that the bat hits the ball, which can be inferred by timing the entry and exit of the bat crossing 4 IR beams and a little geometry(assuming a fast timer and narrow IR beams). For the purpose of my project, the miniscule curve within this square region is negligible. I can assume a straight line between the points and get a good idea of the overall swing plane. So, given that the bat is travelling in a certain direction, (must account for left/right batters), the bat should cross the beams in a certain order and measuring the timing between those crossings approximates the speed and angle. If one or both of the horizontal beams do not trigger, then the plane is fairly flat. If I place the IR transmitter/receiver about 2 meters apart, the bat will cut into the plane and swing out without hitting them.
Attached is a diagram of what I am thinking. Am I oversimplifying or overlooking something?
Thanks again for listening to my ramblings.
This is a pretty good idea! Typically, clever think can easily reduce the hardware required.
Depending on how fast your beam + IR + code can trigger and measure the timing, you may very well meet your requirements using those 4 beams. Worse case, you can simply add a few more beams (at very little cost) to determine more information (if required).
I’m looking forward to see how well this will work!
I found a bunch of Banner DS18AP6LP IR Beam transmitters/receivers on eBay for about $8 each, so I will try using an array of these. The datasheet says they react in about 32ms which is fairly slow, but since the bat has to break multiple beams, all I need is the delta between the crossings, not the absolute start.
I did reach out to OmniPreSense about their OPS242 and they responded right away and over the weekend! In looking over their datasheet, it might be a worthwhile combination to use the IR beams vertically to measure speed and the OPS242 horizontally to track Y axis movement.
Thanks again for your help and once I start start experimenting, I’ll post some results.
As long as your sampling of those is accurate and fast (and the delay in the response/capture is reliably the same), you should have an easy time getting the data you want from it.
Glad to hear that! They are most certainly one of our most responsive manufacturer that we distribute products of.
Eager for the results! 
