A comprehensive overview of the Aero Ranger Cloud dashboard, which allows you to view all of your vehicle & license plate data captured with our EdgeAI ALPR/ANPR systems

Video Transcription:

Hi guys, my name's John and welcome to another video about Aero Ranger.  

In this instance, we're going to look at the Aero Ranger cloud management dashboard - and this is where I think all the magic happens, actually, within the system.

 

A lot of the other videos are about collecting the actual information about vehicles out in the field such as the plate, the make, the model, the colour, the age, that kind of thing, and this is where it all ends up, and it actually ends up here using 4G data or 5G data depending on which device you have. Nice and quick and simple and easy to use.

 

So, let's jump in and I'll show you some of the features that this management dashboard has. So, essentially, it's allowing you to search without looking, a bit of our tagline there. This is our new studio and we have yet to completely sound deaden it so hopefully it's not too echoey and we do have slight traffic noise outside so I apologise about that prior to this video. Well, halfway through the video. Let's jump in and hopefully you can hear me loud and clear.

 

So, I'll just log in, and right at the start we're met with a simple widget that really shows us what kind of volumes that we've been looking at over the last couple of days. So today in the last 24 hours we've had 46,000 vehicles being collected by our cameras out in the field, I suppose, and these are cameras that we've been having out there for a number of years for testing for different types of reliability under weather strain, I suppose, whether wind and rain and very much heat - today was over 40 degrees Celsius in Australia, I forget what that is in Fahrenheit.

 

So, yes there's a lot of cameras out there doing the collection for us and it all ends up here in the cloud management dashboard. So, we've got 46,000 vehicles collected today, eight different categories those vehicles are under and the alerts that have come in from today from those 50,000 is zero, so we haven't actually had any alerts today. Alerts are vehicles that we've put into a hotlist and if they're found anywhere within any of our cameras they will be added to this list here. The reason there's probably nothing in there is we've actually got a very small alert list on this test account, there's no need for us to actually scan and be alerted on lots of cars. But I’ll show you how to add some later in the video.

 

On the right-hand side, we've simply got the most frequent plates. Now, this is normally buses, bus services and things like that which are the 'TP's - they're just going around quite regularly so they're understandably high up in the most frequent plates.

 

We've also got a number of others and that might detect some kind of suspicious activity, I suppose if there's some sort of courier type issue going on. Usually the case though, maybe there just are vans doing their regular rounds. Okay, let's jump into the Plate Search page. Now, this is where most of the work happens I suppose, the hard work, the searching for plates on a regular basis.

 

In the first instance, you can just select the last sort of shortcuts of days and there's a sneak preview coming very soon where under 24 hours there will be a ‘Live’ button as well. So, look out for that in the very near future.

 

So, let's for this example click, let's go with 14 days. So that's actually going to collect the data for 14 days, and it's very quick to actually find half a million vehicles just like that. So, we sort of pride ourselves on, you know, the speed of actually gathering the data, and if we zoom in on the heat map you can see all the vehicles that have been collected over the last 14 days.

 

So, a couple of checkpoints, they're in yellow and this particular checkpoint has found 176,000 vehicles. There should be another one down here, this one here - 360,000 vehicles and roaming vehicles that are sort of dash-mounted to cameras on vehicles that are going around. So, the blue dots are 'reads' - simply cars that are not on any list and they're just sort of being collected for search purposes at a later date, but also different colours will come up if they are categorized as some kind of misdemeanour or even whitelisting scenarios.

 

So here we have three here. So, what's happened here, we can just sort of surmise is there's a car that's been going past and collected a couple of shots of the same vehicle that has been categorized and in the Alert page. So, let's click on this.

 

So, we have quite an old car here - a Datsun 260, green, 1980 to 1984, sports sedan, and the hit details says it's a ‘Quarantine’ vehicle. So, if we go to this one here, same car, same driver, Datsun and 'Quarantine’ vehicle' again. We can actually zoom in and see a bit more detail about this as well. Yeah, so you can really start to gain a lot more information about it.

 

So, let's just close this down and zoom out again and close this off. But you can collect, you can click on any vehicle. Let's click on this one for example, and there we have a Toyota Yaris, Blue, 2010 to 2014, and the full plate there as well.

 

Just to show you how quick this thing is let's go '30 days'. So hopefully we'll just be a couple of seconds and we'll bring up '30 days' just like that. Over a million records.

 

Now if you want to go further then 'More filters' can do that for you. Click that and you can go back as far as you wish. Let's just tax the system and go back to the beginning of the year. So, this is the full year of 2020. Now we're sort of a couple of days before Christmas. Let's see how quick it is. So, I've just clicked 'Done' and it's bringing in all the reads from the whole year. Now this will take a bit longer because obviously there are so many more reads to filter, but there we go, we've got it. So, 6 million, 6.7 million just like that.

 

So, in this sort of example what it's doing is actually loading some of those reads on the map, but it's also loading the points in the background as well. So, when this reaches 100% you will see how many points that it's actually loading on the map all at once. So, we're using a sort of a clustering technology within the map-based tech to really sort of push the limits of mapping as far as they can go in the current standards on browsers.

 

So, let's just wait, it's just there now, and just give it a second... look at that. Okay, so we've got six million points all placed on the map, and if we zoom out further you can see all the heat maps as the dots are coming in. If I zoom down into here you can see how quickly they load up for different jurisdictions as well.

 

Okay so there's a couple of other things that are really cool within the map search system and we've designed it to be as simple to use as possible and sometimes those features are a little hidden, but it keeps the view nice and clean.

 

So, let's go back to '7 days' and just let it load the whole thing. It won't take many seconds to do that, and then I want to type in just like you would in google 'Yellow Mini'. So now what it's doing is it's looking for all the yellow Minis in the last seven days. Let's see what it comes up with. Okay, so it's brought out 21 yellow Minis in the last seven days and here they all are. So, as we scroll through it actually seems like there aren't that many different yellow Minis. It's actually got a Fiat just here, but you know, it's actually brought it up because it's a very similar looking car. So even though it got it correct it's using the algorithms to just sort of putting it in the suggestions for us here.

 

Let's click into one and see some of the details. Let's zoom in on the map on this one. So, it's found at Stirling Street and we click 'More details'. Here he is. And we can zoom in and see the vehicle itself here. Alright, let's type in a plate number

 

So, what I can do is just go across to one of our live streams. So, this is actually on YouTube and you can see the evening shots as well. A nice little one of the Perth CBD in the corner there just for fun, and the vehicles going through this intersection here. So, what it's doing is it's collecting the vehicles on the edge and sending them to the Cloud.

 

So, let's pick out one of these - '1GGL', '1GGL'. Okay, let's pretend I didn't get the whole plate - '1GGL', that's all I know that's in this plate. So, all I need to do is - let's go 'Last 24 hours' so that loads, and I will go one '1GGL', if I don't know what the rests of the plate is, I'm just going to put star as a wild card, click 'Search' and what it's going to do is look for all the plates with '1GGL' in them and then we can start looking through them. It is done chronologically, so this is the vehicle that we just saw on the live stream, so let's zoom in on that map - there we go, and that was the vehicle that was captured. So, if I again click on the details there, here he is, just going past the bus stop. If we go back to the live stream there's the bus stop as you can see, so it's that quick. It actually sends the small packets of data across to the cloud, it doesn't stream all the video footage, that would take up a lot of data it would take up time, it would take up computing power, this is designed to be super lean and very quick to search.

 

Okay, let's get even more specific. Let's go into the type of vehicle, the colour of vehicle,

in a geofence over a certain amount of time. Now, this might take a little longer to do the search because you're actually searching a lot of parameters within the database. However, if you want to drill down into those types of categories and filters and you're willing to wait maybe up to a couple of minutes, then it's well worth doing. Let's show you how that works.

 

So, let's go to this sort of area and use a Polygon Search to select an area of interest. So, what that will do is that will filter this particular area and it is found in the top right-hand corner here. I want to search in the last 14 days, but before I do that I want to type in 'Ferrari'. Is that how you spell Ferrari? Ferrari.

 

Okay, I want to look for Ferraris in that section in the last seven days. Okay, there we go, we found four Ferraris in this section in the last seven days. It's that simple, and it looks like it's the same car every time. There we go, alright, let's go across to a few more features within the Aero Ranger Cloud dashboard.

Let's jump into 'Alerts' very quickly. I won't spend too much time on this because it is a kind of a slightly more boring part of the management dashboard, but very important at the same time. Here you can add in plates and they will be added as an alert so whenever they're found by any of the vehicles or any of the Edge devices going forward, even the mobile phone, it will alert the person that is actually next to the camera at the time, but also send out an email to the people that have subscribed to that particular plate.

 

So, let's just for interest click on this one and edit this one. Ah, I haven't got the accreditation to do so, let's do this one because that's me. Okay so this one is '1GOK179' and then 'Mitsubishi Triton', and I've set the alert for 280 hours. Now I can actually set that for 280 days or 365 days or 12 months, it's up to you, and it's really depending on how much you want that plate in the system for, and it'll auto-expire if you put in '1 day' for example. That's very good for short-term alerts.

 

Here's the message, a message is just a message. That seems like a useless message to me, so let's just put in... and I apologize if that's spelt wrong, I'm terribly dyslexic. Okay so that will actually do the text-to-speech out on the phone when that vehicle is found, it'll actually say "Quarantined vehicle" to the occupant of the vehicle or the person that's actually snapped the vehicle with the phone and it will also send alert via an email to that person as well, or the person that's actually put it in - in this case, me. Then submit for approval and it will be updated.

 

So, the actual alert information category can be selected - we've just got some generic ones in here. Like I said before, let's just go with 'Quarantine' and submit for approval. So that one's been updated and next time that one's found it will come up with the message that we put in.

 

Okay next down, Camera Sources. This is quite a fun page. This shows all the cameras that we have out in the field, and if they're green, they've been reading and sending information in the last four hours. If they're orange they've been sending information in the last 24 hours but nothing less than four hours, and if they're red they haven't sent anything in the last 24 hours. So, it's really good for a status page in terms of are those static cameras up and running and reading, really good for fault finding, and also good for just working out if there are any issues with the cameras and just seeing if anyone has utilized them or not. If they're black, there has never been a read from that particular vehicle or camera.

 

Okay, what we can do now is click on any of these and find information based on stats from that particular camera. Let's just jump into this one here and what this will do is bring out some cool information about that particular camera. So, in the last 24 hours this one, Charles Street North Perth has found 19,000 vehicles and as you can see the peaking troughs of the rush hours as normal, as you would probably expect. A plate found 540 times a day, that's pretty interesting, might be worth a search at a later date just to work out what that plate was doing. The average confidence of the actual make is 75%, but the plate is usually 95% confidence or more.

 

If we zoom down you can see it's been split down into 'Body Type Over Time' as well. So, sedans get the most popular vote for vehicles coming in, and a lot more coming through in the sort of rush hour in the evening. Then SUVs, compacts, and crossovers, trucks and utes as well. So, Utes are sort of busy all the way through the day.

 

If you click off one of these, what it'll do is take that one out of the Search and give you more information. So, you can drill into a bit more information. If you tap on any or hover over any of these, it'll give you the exact figure. So, 72 utes came through at 1 o'clock, or the hour of one. And then this is kind of just a pie chart breaking down the whole day.

 

 Now we've got 'Make Over Time' - as you can see Toyota sort of romp away with it in Perth, especially on this road, and then Mazda comes in 2nd alongside Volkswagen. So, if we take out Toyota out of the mix, then you can start seeing the other vehicle makes & models coming in as well.

 

At the bottom, we've got a nice sort of more standard base list, and this is the very latest list that's coming through, so this is only, let's just look at '20:03', the time is actually now '20:05'. So, when I actually loaded this a couple of minutes ago, these were the latest plates coming through, and that's why they're dark. You can look at them in tile mode as well, and you can also... if we go back here, go full screen, at this point you can export as a . CSV as well.

 

Okay, if I click on '7 days' what it'll do is condense that and bring that out to each day per point and load the last seven days. So here was the weekend, you can see the dip. If I do the last 30 days - let's give it a couple of seconds, and I’ll bring in the last 30 days and you'll be able to see some trends. Here we go. So, it's bringing in some of the information since we actually started it, which was around 28 days ago. So, starting to pick up, probably because of the Christmas traffic towards the end of the month, and a big dip there, maybe it was a public holiday.

 

Okay, so 331,000 vehicles coming through this month from Charles Street. Again, as normal broken up by body type, by make, and here's the plates as well. Cool so that's just one plate, obviously you can jump in - sorry one camera source - you can jump into any of the others and it will do the same thing even with plates and even with mobile phones as well, it'll be able to bring in that information.

 

Just as an example let's click on this BMW. The information will be slightly different because it's not going 24/7. It's only when that vehicle is out and about, but quite good to just see the activity of the particular vehicle. So, it's good for that purpose, if not for getting stats.

 

Okay, other than audit logs which will actually trace from an organization perspective all the activity within this system for administration-eyes only, I suppose, that's really a lot of the features for Aero Ranger Cloud dashboard in a in a nutshell.