Krystian Michalak
JnJ
Nov 18, 2022
Period
- 2021
Title
The Solomon system was tested on-site by vision and robotics engineer. Customer product body lotion pumps were placed below Solomon camera and the system was then started to detect and pick up the pumps. At first, there was difficulty in getting the system to work but after couple of weeks of careful calibration and assistance from the Solomon team themselves, we managed to get the Solomon system to pick the pumps. The camera talked to a UR5 robot in order to pick the pumps up from a bin. The parts in the bin were randomly placed. The challenge of the project was to determine which pumps should be picked up first.
Pro
The support from Solomon was nearly instant, got engineers on site to set up equipment and do training.
The teaching of the objects was user-friendly.
Con
The software was horrible, with version problems - every version had bugs, and the programs made were not compatible between versions.
The UI was very simplistic and initially thought this was a good thing but after a while, it didn't allow us to perform tests we needed to do.
Quite expensive for what it is.
The focus can be tweaked, BUT the camera has to be physically opened destroying the warranty seal.
There were bugs in the software virtually everywhere.
The UI also looks very childish and amateur compared to the same spec same price competitors.
To summarise, the Solomon camera took a long time to fully calibrate and make it work at a level that would satisfy the use cases. The software which would get updates every week would fix issues but also create a lot of different issues. After careful testing and calibration, we discovered that there are simply too many different issues with the system. The Solomon team tried to develop PLC capabilities for our needs but failed to deliver on reliability.