Gap Monitor
Automated Postal Package Sorter (APPS)

My role in the Gap Monitor project was to take the prototype and turn it into a production product. Some of the less mechanical problems I had to solve to implement the gap monitor were: creating the SSH tunnel code to transmit data from the system to the database, add a standard IP address to the system,  wire the system into the postal network by routing it through the machine and ensuring the length was not too long, and create a kernel level code on the system to start the monitoring program when the machine powered-on.

 

There is a second type of bolt on the machine, an expansion bolt, we were able to capture it fouling within the first two months of deployment. Which is special because they never fail. It allowed us to show how we can use the system to monitor bolt health. 

 

At this point, I started working with a vendor to create a beta version of the monitor that could be mass produced for national deployment. This included providing copies of current cad files, sensor models, and the computer system. I reviewed and signed off on design changes which included standardizing the chanel sensors and revising the upright brackets to be easier to mount and have breakaways. 

 

We then tested the two systems against each other to ensure the vendors worked as well as the in house version. Speaking on data acquisition they were comparable. However, the vendor's system had an issue with durability. Their board fried after about 3 months of being hooked up to the machine while our version still worked as intended.

 

Another aspect of this project was learning how the system might fail and how those failures can be minimized or addressed. Two examples I documented and created maintenance routes for are blocked sensors or misaligned sensors, both of which will throw a no read error. The solution was to ensure ops took a quick look at the sensors daily to check that they were aligned and not blocked.