As greater numbers of OpenROVs are getting built, we're getting more data on the kinds of technical challenges people encounter during construction and operation. As the various causes for anomalies are figured out, we're also getting data on how those problems can be solved. Well understood failures are an excellent source of data, so I thought it would be good to start organizing common failures modes and their causes. In the short term, these data can be used for simple troubleshooting, but as the network of failures and causes gets more complicated, I think we'll also start being able to construct an "anomaly model" that gives us an overarching idea of the vulnerabilities of the ROV and ways to manage and prevent failure in general.
As an example, I've constructed a (still quite incomplete) flow chart of possible causes for one specific failure- namely, not being able to connect to OpenROV Cockpit. I'm sure that even this tree of problems and solutions can be greatly expanded based on observations we have all made so far. My dream is that we'd be able to have a sort of 'living anomaly management flow chart' that everyone can contribute to as problems and solutions are identified.
By the way, if you already have a specific modification to this flow chart, you can edit it as a Google Doc here (you'll have to install lucidchart first). Also, if anyone can suggest a better cloud-based, corroboration-friendly flow chart system to use, let me know!
Added my two cents on BBB lights not working! Looks like my unit got the power IC u2 on the BBB destroyed. Either because I didn't glue the ends of the expansion IO wires or because I didn't seal something right before putting it in my bathtub for buoyancy adjustments for goPro.
I think there is a lot of interest in knowing test results as well as failure modes and workarounds. People in the community have expressed a strong interest. Those considering building an openrov (like myself) could be put-off by hard it can be to find pressure test results and failure modes. I bought my kit knowing that depth hadn't been thoroughly tested yet (now it has been) and there were lots of warnings about leaks and workarounds on the forum.
I would suggest publishing a tests/specs and models summary page. First start with a new section on the GitHub with excel spreadsheets summarizing (loving) all your pressure tests to date. Then worry about how to make the data look nice on the website later.
A. Summary of Depth Ratings by Model (American Bureau of Shipping type: tested x1.25 for two cycles without failure.)
B. History of Pressure Tests & Failure Modes
C. Summarized Scientific Analysis from Forum (fluid dynamics results, amp draw measurements, thrust measurements.
Example with fake data:
Electronics tube and battery tubes pressure tested in test chamber for 1 hour. The most frequent failure mode for max depth was the electronics end cap at the rim. The most common failure mode for leakage was found to be (battery or electronics housing) ...
Model Depth Failure Mode Failed Parameter Est. Depth Rating
2.5 100'x1 (failed) - EndCap@Rim, EndCap@Rim 3mm thick 100'/1.25 = 80m (max limit)
2.5 90'x2 (passed) - NA EndCap@Rim 3mm 90'/1.25 = 72m (safe)
2.5.b 125'x2 (passed) - NA EndCap@Rim 6mm (glued double layer) 100' (safe)
I don'tknow how to add it to the chart but I have a possible connection issue to look for.
If the topside adapter has a connection, but the rov adapter does not,
And, the topside adapter is powered by a wall outlet, even if it is through a laptop or desktop,
And, you (or possibly neighbors?) have homeplug adapters in use on your electrical grid,
Then, isolate your adapters electrically from the grid. Unplug your laptop and run on battery power and use the rov batteries to power it instead of a power supply.
I had a lot of connection issues when I was trying to connect through 2 connections at a time. They can connect through any of the wires and in the strangest ways.
You can Log into them and change their network names. I haven't tried this yet, but the instructions say it will keep them from connecting to any other network. The same instructions say the adapters will connect to any other adapters. Syncing them with the buttons didn't really help.