Tasemu

Afternoon, the app just got through the approval process and i've sent out a link to all testers. Its bare bones but hopefully enough to play around with and give feedback. Check your junk folder if you can't find the email. 👍 I gotta agree with adding the security checks and in app photo taking. I've put it in the list

Currently got about 20 from here and my facebook group I imagine i'd probably stop at around 30-40 for the first run.

Yeah totally get what you were proposing and it would be awesome, i do hope that it has nothing to do with me though haha. And you're right you could totally use an LLM to to match boat photos and places. I do work with LLM's in my day to day though haven't been involved in the actual training of said models. From my experience interfacing with these models i agree that it would be viable to have an LLM recognise your boat and the area it is in, but not so viable for an LLM to create a fake image that would be visually convincing for a particular boat in a particular place. I think i had got your previous response backwards in my head and thought you meant a black market using LLM's to generate fake pictures of your boat in certain places on the canal and then injecting geolocation meta tags into the image.

Regarding this, i'm not entirely sure how a counterfeit market could operate assuming that the picture of the boat with the number is actually included. As the picture in question would be handled by the local crt representatives for the area, if it was a picture of the boat somewhere else with spoofed geodata, then they should be able to tell this easily from a brief visual inspection of the image. For example when someone sends me a picture of their boat on the K&A anywhere between devizes and bath, i can tell where that boat is from the surroundings? And if i were CRT and felt as though the image was too vague to be able to recognise the area, and i was suspicious of the person in question already then i'd just deny the picture. In the end its CRT's decision at the end of the day, this is just supporting evidence that might help. I've used geotagged images to avoid a 6 month licence once before, and it was just a simple mistake with the spotters and all worked out fine.

Not sure if this helps at all, but here is a PDF report i recently exported from the app. Its early stages at the moment but you can see i've made sure to at least try to ensure my boat is recognisable, with the number showing wherever possible and in an area that the local spotters will recognise matches to the geolocation data presented. bea91ef6-4a3d-49d7-9407-c21036d6f37d.pdf

Yeah i was thinking the same, if i were CRT and someone gave me a picture of an empty bit of canal without the boat and number registered to their account in the actual picture, i'd be very confused.

It currently only allows photos to be logged one at a time, but i can't see why i couldn't add that in. The current bottleneck is that i need to filter out any photos without geotag data attached to stop people getting loads of errors when they select pictures with no lat/lng. This means i can't open the native gallery and have had to make my own version inbuilt. Food for thought 🤔

Ominous... I can't say i totally agree. But i see where you might make the connection. What makes you so confident that a system designed for new drivers at high risk of claiming on their insurance policies would be implemented for a licence system for barged travelling through the uk waterways where congestion is the main cause of concern?

The GIS data for navigable canals and rivers are publicly available from the CRT themselves. I used this data to calculate the distance along the line between the two furthest points in the picture array.

Wouldn't be against this in the slightest, though i'd struggle to get it into v1.0 Nope, it only has access to the photos you choose the log into the app and then uses them to generate the markers and stats.

Currently it only accepts pictures from the gallery, i haven't had the chance to build anything like an inbuilt camera functionality yet, but its already on the todo list. I will definitely look into how these apps are doing this when i get onto building that part out.

Currently there is an option to generate reports yes, they can be shared to whomever one likes. Its a little rudimentary at the moment but fully functional. Also multiple boats is definitely on the todo board.

This reminds me of a thought i had earlier in development, one of my mates dropped his phone in the river by accident and thats what made me think i should be backing the data up online so its not lost in the event of a phone loss... But perhaps those who are more sensitive about their data might prefer an option that stores everything on the phone..?

Currently i store the data on firebase, i see no major issue with this though nor do i scoff at people who might. I try to use the bare minimum required information to make the app work, emails don't need to be verified. I'm not sure what google or anyone else would want with geotagged boat pictures, and i imagine if anyone has online backups on their phone, like many smart phones do, they probably already have them. If there is a real desire for it i could beef up the encryption... I just don't see how valuable the data would realistically be. 🤔

I don't think i'd want that. I already have a job, not after a second one.

This is all done already, but you're right it can be faked. However I believe that this could prove a useful tool and i've personally seen many instances of CRT disputing a CC'ers journey and then accepting proof that was magnitudes of times less thorough than this. Also i believe this to be a tool useful for sorting out misunderstandings, not something to try to use to force CRT to take your word for it. If they don't trust you or think you're bullshitting and you are, this wont save you. 😆

I totally agree actually, this is an option tool for those who want to be able to back it up if they know they're doing their distance. The idea has come from a few friends who got stung because the spotters were out as they were moving, or there was a mix up. I've had it once myself and was lucky i took a picture at the time.

My thought process was by collating a few things, one being the actual picture of the boat with the licence number showing, ideally in a recognisable area. Then it would take the metadata which is attached to the picture at the time it was taken. The latitude, longitude and the date. So if CRT for example wanted to verify that the boat was indeed there, they could visually match the area in the photo to the lat/lng supplied.

Hi all, been a while since I've posted here and i've decided to take a break from my usual begging for engine help to sit down and write something that I believe might help the boater community. I posted this on my Kennet and Avon facebook group earlier but thought it might be prudent to post it up here too. I wrote a snazzy intro i've copied below, but anyway I hope some of you might find it interesting. --- Hello everyone! 👋 As a solo software developer and fellow boater, I've been working on a solution that I believe could help us all: an app called Drift Diary. Drift Diary is designed to track the distance we travel on our beloved rivers and canals over the course of a license year by allowing us to log photos with geotagging data, creating a digital diary of our journey. Even better, it can generate PDF reports to verify travel distance with the Canal & River Trust should they decide that we haven't gone far enough, or missed us while logging boat numbers. But to make this app truly effective and user-friendly, I need your help. I am seeking volunteers from this group to participate in the first round of closed alpha testing for Drift Diary. This would involve using the app and providing feedback on any issues you encounter and your overall experience. Your privacy and data are important, so rest assured, the data collected is yours and yours only, also this app will be, and will remain free for all. It would mean a lot to have your input on this project. I believe together we can shape Drift Diary into a tool that genuinely benefits our community. If you're interested in participating or want to learn more, please let me know! The app has been submitted to the Google Play Store for review for a closed testing run and hopefully should be approved by the end of the week. If you are interested in testing this and own an Android phone then please DM me your email so I can add you to the list, once the app is approved then an email should be sent to you and you will be able to install the app. There are plans to support iphones too in the future, but Apple make this prohibitively hard/expensive for me to do at this time. Thanks in advance for your time and consideration. Happy cruising! God those are some huge screen-shots... 😕

Thanks! Would you agree that the above symptoms mean a borked AVR? Would be a shame to buy a new one just to have it do the same thing.

Ok so i finally have an update for this thread. Through much nightmare I was able to disassemble the the acoustic box from around the generator and gain access to the end cap, which when removed allows access to the AVR and brush assembly. before changing anything I did the following tests with the following results. 1: Ran the generator as-is and off-load, then measured the Hz / Volts. Result was 45hz @ 260v. 2: Adjusted the engine speed carefully to bring the Hz up to 50.8hz off-load. Resulting voltage was 290V. 3: stopped the engine and adjusted the AVR voltage adjustment potentiometer down a half turn before re-starting. Result was 290v. 4: stopped the engine and adjusted the AVR voltage adjustment potentiometer to its minimum value (full ani-clockwise) before re-starting. Result was still 290v. Notes: Brushes appeared to be in decent condition with enough material left to make good contact. One of the slip rings was slightly worn to the touch but didn't appear to be significant. Reading previous posts on here, i'm tempted to guess that my AVR might be busted and a new one required. The AVR is a Markon AVR 036-245. I'm not sure if it is still possible to get one of these, or maybe there are other models that might fit? Thanks in advance for any help!

Excellent, this clicks for me. Thanks!

I see... So am I right in my understanding that this doesn't actually cause any issues with the charging, but just with the warning light? or will the fact that only 2/3 of the field windings are powering the rotor result in a 1/3 decrease in current to the batteries (but still a good voltage)?

Could you explain your reasoning to me? I'd be really keen to understand and learn

We connected d+ and b+ using a multimeter. Upon starting the engine we were getting a voltage drop of 10.5v, when increasing rpm to high speed the voltage drop went down to 0.3v. when returning throttle to idle, the voltage drop settled at 3.5v. also the indicator light dimmed and brightened accordingly. We used little alligator clips, cause its so damn easy to short against the casing by accident! EDIT: noticed we didn't supply a reading at moderate rpm, so here as an amendment to include it: 2.2v on moderate revs. 0.3v on very high revs. 3.5v on returning to idle