Ring Package Analytics Tested
More sophisticated analytics have the potential to solve major problems, such as package theft. Recently, Amazon's Ring has released package detection analytics for their doorbells, intended to alert users when they receive a delivery, but how well do they work?
We tested Ring's package analytics for over a week to see where it worked and where it did not, examining:
- How accurately did the analytic detect packages day and night?
- How did package size affect performance?
- How did distance affect performance?
- Did the analytic detect other objects as packages?
- How we recommend these analytics can be improved?
Related, for another recent consumer analytics development, see Wyze AI Smart Vision / "Anything Recognition" Tested.
Package ****** ************
******* ****** *** ********* ** **** doorbells ******** ** *** **** ***** Doorbell (*** **********) *** **. ** is ******** ** **** ******* ***** ($3.99/mo ** $**.**/**** *** ******) ** Ring ******* *** ($**/** ** $***/**** per ****)*************.
******* ****** ** ********** **** *******-******* doorbells ** **** ** *********, *** Ring ***** **** ***** ******* ** motion ****** **** ***** *** ******* faster. ** *** *******, ****** ******** ***** ~**% ** ******* **** ** 10 ****, **** **** ******* ** alerts *** ** ********* *********** ** testing.
**** ********* ***** ****$**.** *** *** ***** ******** ***** to $***.** *** *** ***** ******** Elite (***).
Competitive *******
***** ******-**-******** *******/********* **** ******* ******* detection, ******** ** *** **** **-** months, ************* (** **** ** ***** **** Cam **** ****)********* **** (**** ** **** *****). ******, **** ** ****, *****, Eufy, ***., ** *** ***** **, though ****** ***.
*******
** **** * **** ** *******, Ring's ******* ********* ****** ** ****** most ******** **** *** ** ****** in *** **** *****, ********* **** large ******** **** * *** **** from *** ******, *** ***** ******* alerts **** ******** ** ****** ** multiple *********.
***** ****** ********* ********** ** ******* and ****** ********* *********** ********, **** performance ** ****** *** ** *** than ** ******* (******).
Recommend ************
***** ********* ***** ** *** **** within *** ************ ** ***** ************ analytics, ** ********* *** ********* ************ for ****:
- ******** ** ***** *******: ***** *** foremost, ***** ********* ***** ******* ******* from ******** ** * ******* ********* of ********. *********, **** ***** *********** packages *** ******** ********, ***** ***** boxes, ********* ****** ******.*** *** *****, are *******, ** *** ****** ***** and ****** *********.
- ****** **** * ******* ** *******: Secondarily, ********* ***** **** * ******* has **** ******* ***** ** * useful ******** ** ******* ** ******* package *****. *********, ************* *** **** sent **** *** ******* ** ***** detected.
- ***** ******** ***: *******, ** ********* that **** *** * ***** ******** FOV ** ******** **** ******** ** front ** *** **** *** ** detected. ** *** *******, ***** *** a ****** *' ***** **** ******** below *** ******, ********* ** ****** detection.
Does *** ****** ******* *******
****'* ******* ****** **** *** ****** packages ***** *******, * *** ******* in *** **** *** ***** ********* ***** *** *********. **'* ******** **** *** ******** may ****** * ****** ******** * package ****, ** ** ******** ******* packages ** *** ******* (********* *****), but ** **** *** ************ ***** on *** *****/******* ******.
Package ********* ******** *****
****'* ************* ********** *******, ****** **** ** ***** best **** ****** ** ***** *****, with ** ***** ************ *** ********, perspective, ** ***** **************:
******* ****** **** **** **** ******-**-***** shipping *****. *********, *****, ***** *****, and **** ********** *** *** ** detected. **** ******* **** ****** ** Motion *********, ** **** **** *** have **** ******* ****** **.
** ****** **, ***** **** **** it ** *** **** * ******* detection ****, **** ******** * ****** zone, ******** *****.
Package ***** ********
** ****** **** ******* ****** ***** default *** ******* ******** *** ***** no ********* ********** ** *********** ** either *****. ******* ******* *********** ** Max ******** ** ******** **** ****** on ******* ********, *** ** *** no ***** **********.
********** *********** *** *** **** ************** Ring **** ******* ********* **** ** asked ***** ****** **********.
Test ****** ** ********
** ****** ***** *** ********* ******* Zone ************** *** ***** ** ********** in ***********.
*****, ** ******* *** ** *** Package **** ** *** ***** *****, so ******* ** *** ****/******** ***** be *******.
*******, ** **** ****** **** * taller *** ** ******* ****** ******** when ***** ** ******* *** ***** or ** *** **** ** *** porch, ****** **** *** *** ******* more ******** ******.
**** ******* **** ** **** **** drawing *** **** ** *** ****** should **** ******* ******.
Detection **** ** ***** *****
** *** *******, ******** **** **** detected **** ******** ** ***** ** the ********. *** **** *** ** shown *****, **** ******** ** *** left ********, ***** ***** ** *** other *** ** *** ***** ~**' away **** ***.
*******, *** ********* *** *** ********** even ** **** *****, **** *** analytics ****** ******** **** ******** ** close ***** ** **** ****.
Doorstep *** ** ***
***** *** **** ********'* ********** *** is **** **** (***°), *** ******** AOV ** **** **°, ** *** area ******** ** ***** ** *** door ***** **** ******** *** **** is *** ** ****. **** ***** a ***** *** ** ***** *** camera **** ** ********** *** ****, but ** ******* *** ** **** even **** **** *** *********.
** ***** *****, **** *** ****** of * *** ****** ****** ~*' of *** **** ** ******* (*** not ******** ** *** ********).
** ********, **** *********** ***** **** taller ****, **** ** *** **** doorbell ***** *****, ***** ******* ******** the ******** *** * ******* **** on *** *******.
Missed ******* **********
****'* ********* ****** ******** ******** **** in *** **** ***** **** ****** deliveries *** ******** ******** ****** **** testing.
** **** *******, * ~*' **** package **** ~*' **** *** ******, along **** * ****** ******* *** were *******:
******** ** ****** ********* **** ***** detected. *** *******, *** ***** *** in **** ********** *** ******* **** delivered *** ***** ****** ** ******* times.
Package ********* **** *******
**** ********** ******** ******** **** ******* very ***** ** *** ******, **** as *** *********** ******** *** **** below:
*******, ******* ************* ** ***** ****** are *** **** ****** **** ******** detected *****, ** ***** *** ****** to ****** ** ***** **** * person ** ***** ***** ****** **********.
Package ********** ** ******
******* ***** ****** *******, *** ********* produced ***** ******* ************* **** ***** subjects **** ** *** *****, **** if **** **** *** ******** *** packages.
Versions ****
**** ***** ******** * ******** *** "Up ** ****" ** *** **** of *******:
******* ******** ** ****** ******* **** Ring ******* *** **** *** **** Camect. **** ** *** * ***** on ****** *** *** ******* ** Ring's *****. ** *** ********** ******* on *** ********* ** ******.
**** ** ***********. * *** ******* on **** **** *** * ***'* read ** ******* ** *** *** responses.
****'* **** *** ***** ******* *** why *** ********* ***** ****** ***: ******** ******* AI *** **** ******** ********** **** ****** **** ***, *****, Amazon, *** **** ****** *** ******** are ********?