Go drive/walk by your local police station and note all the plates of unmarked cars in the area. Many of them are likely police vehicles or private vehicles of employees. Do the same near a courthouse. Use those as fodder for this tool.

Granted it’s been 30-something years since I graduated with my CS degree, but my experience might be helpful. My first job out of college ended up being in tech support for a software company, and I used that to eventually move into software development. During my 2 years in tech support I learned what I could about the source code for the companies product, and built relationships with members of the engineering team. It got to the point where I could carry on conversations with some of the software developers to identify potential causes of bugs that customers were running into.

Because of the relationships I made with the software engineers they eventually recognized my skills and one of the managers eventually helped me transition from tech support into a software developer position.

So start feeding it lots of random 2 to 8 letter/number strings.

I don’t know if this incident involved Flock or a competing product, but John Oliver recently described an incident on his show that demonstrates one of the major issues with this technology.

Apparently an ALPR reported a “hit” on the license plate of an SUV in a shopping center parking lot. The vehicle was reported stolen. So a half dozen or so police performed a felony arrest of the woman in the car along with a number of family members, including small children.

The problem was that while the letters & numbers on the license plate matched, the actual stolen vehicle was a motorcycle from a completely different state. So if any of those half-dozen cops had bothered to confirm even one additional detail about the stolen vehicle then this wouldn’t have happened.

How will Flock improve their technology so that police actually do more than the bare minimum when being told of an ALPR “hit”?