I use a ton of different types. I think my favorite is the EM deflection sensor, which is normally advertised as a "microwave radar" (it's not really either) sensor:
The output is binary so it's easy to wire up. The sensor works through non-metallic surfaces so I can put it inside work boxes and still detect movement nearby. It's really fun to make Halloween jump scares with it since it can be totally invisible to people.
I'm currently starting to use LiDAR sensors. It's more work to do the background subtraction, clustering, and interpretation, but it's coming together. I want to make skeleton heads on servos that track people coming up my walkway.
For now, I'm working with Slamtec's RPlidar A1. It's cheap, does 360 degrees, and has basic libraries available: https://www.slamtec.com/en/Lidar/A1/
Super cool! Lidar is definitely the future in a lot of ways it's very useful. Do you have little microwave emitters that you use to detect when the signal is cut, like an invisible trap wire?
I'm gonna have to check out this chip once I get home.
I’ve been working on a project of a digital bubble level with gyro stabilization. I’ve done some fairly intensive testing with different gyro/accelerometer combo systems. Of of the lot, the BNO085 really stands out because it’s the only one (apart from the BNO055, which is the previous version) that does the fusion of the two sensors internally and output a AHRS (angle of the sensor position in the world) without manually implementing a complicated algorithm library (out of all the libraries I tried, the Fusion aka Madgwick2 was the best but still a real pain to implement). It’s designed for VR/AR headsets and gaming, the down side is that the whole match is black boxed so it’s harder to troubleshoot if needed.
Cool! That would definitely be a huge time saver if it works properly. How has the project come along so far? Do you have any pictures to share?
I'm also curious how responsive the chip is, and if it can refresh often without corrupting it's data output? High quality, all in one gyro/accelerometers are definitely gonna be needed in the future of robotics and so I'm gonna have to check out the documentation for this sensor before I start my next project.
My last iteration was actually with a Arduino Nano rp2040 which acc/gyro on the board and I spent way too long to try to align them properly (it’s too bad the board had all I needed on one PCB) now I’m back tracking to the BNO sensor. It’s pretty fast and simply sends 3 coordinate values at each cycle. So you just have to check for fresh values and read them at each loop. The board basically has its own MCU on BNO board for the math so it never saturates your clock cycle.