New uploads!

New photos have been uploaded to our gallery, bringing the total photo count up to 700+ now! Due to a camera crash, we did lose about a day and a half worth of photos, but the crash has been resolved and was happening because the SD card had gotten full again.

Photo gallery up!

We upgraded our hosting to the top tier offered by our host. We should have plenty of disk space to host a ton of photos generated by this project over the course of its life cycle. And as far as needing our users to mirror things such as time lapses, no long an issue, as we have unlimited bandwidth on our plan now.

With this being said, our project is proud to announce the newest addition to our project, the photo gallery that is being offered at https://gallery.barkshark.net! Also, it is in the works currently, but we hoping to get our media director setup with an email address to allow outside projects to email us pictures made from the data or setups for us to host. Until we get her email setup and working, that will be something to come down the road.

HTTPS Enabled

I found out that our webhost generates an SSL cert from “Let’s Encrypt” and have enabled HTTPS to encrypt some of the website’s traffic. Figured this was best since a small bit of the data we handle could be viewed as sensitive information.

Woo! Milestone reached!

So, if you’ve been keeping track of things, you know that we recently did some upgrades to the power supply! We upgraded to a new 50 watt solar panel, a new PWM charge controller and a marine deep cycle battery.

We have been running a test stream for over 37 hours straight now, meaning there have been no resets under the new beefed up power system!

SkyCam Picture – 04/12/19

Seems the camera is getting rained on a bit, wonder what all it will capture today. The roof/ground mount is deployed out in the yard, with the camera setup to make 100 second video clips on any sudden noise and to send a picture every minute to my PC using FTP over the mesh network.
Looks rather spooky, but with the exposure and image gain turned down a little, you can see the cloud formations as they move over our area. Still getting rained on, you can hear the droplet hitting the camera’s face on the audio. (The camera does run a media server that anyone connected to the mesh network can access!)
It actually starting to get fairly dark outside, might have a small storm moving in or just a really big rain cloud. Not sure since I haven’t checked the weather radar today.
I will have to say, the camera getting wet does take away from the imagery a little bit, but what can you really do about such things.
Hmm, seems the little storm cloud has decided to leave for elsewhere.

Additional overflow pictures, seems the storm cloud didn’t like being called small and has come back.

Roof/Ground Prototype in Testing

After much back and forth with the grid paper and tossing ideas around, we finally have put together a prototype for roof or ground based operations. The setup is easier to maintain and make changes to, as it is no longer so close to the ground. We also got the radio up around 14ft or so, which is the biggest part of a mounting system. There are some photos below to show off what we came up with.

Also, the 12Ah SLA battery that we originally intended to use has proven to not be a viable option. Either the SLA battery just doesn’t have the reserve needed to hold 450ma over night, or the battery has a defect. So, we switched over to using a marine boat battery for power. Should be able to hold the system for a few days of cloud cover.

Possible Committee Meeting

Good afternoon everyone,
We met with the Park Manager for Huntsville and presented our project with its goals and current structure. After discussing things such as how the project is powered, making sure the project doesn’t capture photos of passerby’s and so on; I think it is safe to say that we at least have a foot in the door.

After going over some documentation and explaining various aspects of the camera node and its requirements, the Park Manager (Colleen MacDonald) has brought up a point of possibly having us attend a town committee meeting to showcase the project to other folks involved in the town’s operations. In reaction to that, we have began redesigning our project to fit into a small and easier to move PVC frame. We went to Home Depot and purchased all the materials to build the frame, as well as made a stop at Canadian Tire to purchase a higher output solar panel and better charge controller. Up next would be purchasing a better battery.

Oh yea!

We finally located one of the main flaws in our power system!

After much hunting through the system and troubleshooting, it was discovered that the small heatsink on the LM7805 regulator for the camera was not enough to keep the regulator from entering thermal shutdown when the camera was placed into 1080p operation. After moving the regulator to a much larger heatsink, it seems that the random resets have been cured. As we still have not had enough sunny days, we are on the fence about purchasing a bigger battery for the node. We are still going too low in voltage on the battery within a few hours after dark. If we get a sunny day, we will be better able to evaluate if we actually need a bigger battery. My conclusion so far is that it will be a needed improvement because the current battery can only provide nine hours of power in no/little sun time periods. Good enough for a night’s power, not good enough for a few days cloudyness.

Some Link Simulations

In order to plan out the best places to place network nodes, we have been running some link simulations using https://airlink.ui.com/#/ptp. These simulations make the assumptions that both radios involved in the link are placed on a 15ft mast. (e.g., 15ft above ground level for all links)

We are using Lion’s Lookout as the main node point for these simulations and have tested what should be expected at various places using an EIRP of 30dBm. Results are color coded as to how the software thinks each link location would perform

Take this data with a grain of salt. It is a computer simulation that does not account for trees, only topology is accounted for. Real results could be completely different from what is expected. Also, in real world testing, there will be other radios in our path that could cause signal degrading and other negative effects to our communications.

This simulation doesn’t take into account that as radios are added to our network, they themselves become network “towers” so to say and expand the coverage of the network. Each radio keep track of who is can hear and who it can reach via other radios, which this software cannot simulate.

—GOOD LOCATIONS—

Family Place – Signal -53Dbm | Link Capacity 84Mbps | Distance 372 Meters

KFC – Signal -55dBm | Link Capacity 84Mbps | Distance 472 Meters

Mc Donalds – Signal -57dBm | Link Capacity 84Mbps | Distance 576 Meters

Fresh Co – Signal -57dBm | Link Capacity 84Mbps | Distance 602 Meters

Shoppers Drug Mart – Signal -57dBm | Link Capacity 84Mbps | Distance 614 Meters

Huntsville Place Mall – Signal -60dBm | Link Capacity 84Mbps | Distance 812 Meters

Great Canadian DS – Signal -64dBm | Link Capacity 76Mbps | Distance 1.29KM

—QUESTIONABLE LOCATIONS—

Staples/LCBO – Signal -61dBm | Link Capacity 76Mbps | Distance 1.03KM

Farmer’s Daughter – Signal -64dBm | Link Capacity 76Mbps | Distance 1.4KM

Kawartha Dairy – Signal -65dBm | Link Capacity 76Mbps | Distance 1.58KM

Water Plant – Signal -68dBm | Link Capacity 76Mbps | Distance 2.15KM

—BAD LOCATIONS—
Canadian Tire – Signal Blocked – No Link – Distance 897 Meters
CT has a hill behind them that blocks the radio link’s line of sight, making communications not able to happen. Perhaps if CT was to allow our project to place a solar operated node on the roof, it could fill in that area, but that is a far shot.

Poking Town Hall Again, and Test Results

After turning in our first draft of paperwork on the project roughly a week ago, we have decided the first round of testing puts us in a spot that we should be nearing the time to build the first node into its finalized form and start field testing it.

If the field testing goes as expected, the node might actually be getting close to being able to be placed into operation in the town, once we are given permission.

The makeshift test on the test bench proved to hold up a link speed of 120Mbps+ at roughly 700 feet using around half of the nodes possible power output, this isn’t all that far, but it looks promising for us to deploy an open research network based on this technology. (700ft is 0.2km and is the max testing range we can obtain on our property; the radio claims around 5km line of sight, so I’d say 2-3km is doable, but at roughly half the speed…)

To bring things in to the light, we have sent an email to the council on the town’s website and will be awaiting their response. Hopefully they are willing to work with us on this. But keep in mind, even if they do agree to us putting up the first node, it may be a few weeks before we are able to do so. I checked on the conditions of the lookout today and there is no way anyone is getting up there right now, too much ice.

Speaking of the ice, until the weather gets more friendly to our project, it appears that the node will only be able to operate during the day anyway; further delaying some of our testing. Those SLA batteries just don’t do this 12-20’F weather too well and refuse to recharge in those temps. To be fair, the battery is rated for a temp of 40’F minimum.