Why IR OA on the sides?

Joined
Jul 4, 2016
Messages
1,125
Reaction score
799
Age
60
Location
Sterling Heights, MI
Been thinking about it... why? Until it's tested and shown otherwise, I believe that the ultrasonic visual front (and now rear) OA would work better than IR (is that wall much different temp than ambient?)
 
That's how a lot of house robotics like the vacuum cleaners get around obstacles and u can also set up invisible perimeters with infrared sensors that use batteries. From what I understand the Irobot uses both just like our drones. I've had one for years and am still amazed every time it turns on by itself and marches on like a soldier


Sent from my iPhone using PhantomPilots
 
My guess would be cost (maybe) and to cut down on interference. With ultrasonic sensors going off in each direction, there is a huge potential for false readings. I cant think of any other significant reason. The weight would be negligible, and even the cost is not substantial when you are talking about a $1500 drone. Interference/echos from other sensors would be my main guess!
 
Been thinking about it... why? Until it's tested and shown otherwise, I believe that the ultrasonic visual front (and now rear) OA would work better than IR (is that wall much different temp than ambient?)

That's about it BG,
Sonics could be unreliable horizontally due 'looking thru' the prop-wash and related acoustic noise, etc.
Ultrasonic Sensor Operation on a Quadcopter
 
I've only crashed twice and in both cases it was a hit from behind. Once I backed up and wasn't paying attention, was going so slow too! The second time I had swooped into a shoot then whipped her around 180 for the dramatic reverse angle and I didn't compensate enough for the remaining backwards drift. Both times backwards OA would have saved me.


Sent from my iPad using PhantomPilots
 
Off topic but close: has anyone tested, full throttle forward "into" a wall?
Assuming good conditions (right brightness and nice big solid wall), does the P4 really fly along at 30 mph and then hit the brakes soon and hard enough to not hit a wall?


Sent from my iPad using PhantomPilots
 
Off topic but close: has anyone tested, full throttle forward "into" a wall?
Assuming good conditions (right brightness and nice big solid wall), does the P4 really fly along at 30 mph and then hit the brakes soon and hard enough to not hit a wall?


Sent from my iPad using PhantomPilots
Today with OA active my P4 will only do 15 to 17mph, it's nowhere close to the spec of 22mph. I have never figured that out.

For the few in the US that have a P4P today, I highly doubt if anyone wants to test anti-collision at 30mph, but my suspicion is full speed will be more like 25mph with OA enabled, given the over rated P4 OA speed.

As for IR on the sides, I suspect the distance capability is not as far, and deemed not as important sideways for high speed OA. And I would expect the P4P sideways speed will be throttled down slower than 25mph when OA is enabled. Forward and reverse speeds are usually higher than sideways motion speeds, so I assume the IR is "good enough" for a reduced sideways speed of say..... 20mph, and it's cheaper/lighter.

It's interesting to note the Mavic boasts about 24 compute units. I saw nothing about that with the P4P release, stating the number of compute units. Hmmm.
 
Using cameras for obstacle avoidance is VERY processor intensive so my guess is the side will not so stereo optical sensors (cameras) for some time. Sonar and IR may not be any lighter but the processing requirements are typically much less.


Brian
 

Recent Posts

Members online

No members online now.

Forum statistics

Threads
143,066
Messages
1,467,357
Members
104,935
Latest member
Pauos31