Re: [LibreRouter] Access to open hardware?

Thanks Nicolas, we did a lot of work in last two years. We have deployed wifi network in refugee camp that rotated 4000 people each 24h and in 4 months we served over 500,000 people with wifi. Even getting a licence to enter a refugee camp from ministry of internal affairs was hard enough, then we has only 72h to build and deploy hotspots on light poles that has power only during the night. So in two days without almost no sleep we build charging and monitoring system with batteries under 100€ that worked from first try. And it worked for months in harsh outdoor environment. And on the third day we had all of our hotspots deplyed, thanks to lot of people in our local hackerspace that helped and thanks to awesome Nodewatcher system that enabled us to configure all devices and build wifi mesh network in just few hours. I hired professional mountain climber to teach me and help me take all these devices up to 30m high poles and not die while doing it :) I didn't want to risk any of the other guys from our hackerspace, this was my duty to do on my own. What I saw when working with biggest NGOs in the field that none of them are ready for rebuilding internet infrastructure when it goes down, and we have the technology that can be used to rebuild it much, much faster, cheaper and the most importantly at scale! All other solutions don't scale and are insanely much more expensive. I have reseached this field very thoroughly and have talked with all stake holders, like Disaster Tech Lab, communication sans frontières, Vodafon, Unicef, Red Cross, Green peace and lots and lots of small NGOs. And none of then have a solution that scales and is affordable. After the refugee crisis has passed we continued working on MeshPoint. We also build 3 generation of prototypes, did who knows how much case designs and tested lots of boards and routers until we our current version. Our current version works amazingly, to he honest much better that I expected it to work! This is mostly due to modular design which offloads traffic to three SoC so even if one if overloaded we still have two other ones. Also using sector antennas makes a huuuge difference because you have spatial isolation which increases number of users at least 3 fold. Also for conserving power we can always turn off any board if necessary. So we did our first controlled real world test this summer - and we got up to 150 clients connected per radio, so that is total of 450 clients per MeshPoint! We were blown away and after we got these results. Previously our best case predictions were up to 300 clients. But great work that Dave Taht, Toke Høiland-Jørgensen and others did on bufferbloat and make-wifi-faster project payed off huge for us! We are eternally grateful for their work and all other OpenWrt and Lede developers. Our latest tests are deploying tethered drones to establish long range wifi link (check out our Hackaday page for video). We are also testing big tethered balloons filled with helium also for establishing long range wifi links. We are now in last stages of testing open source mppt solar charger for LiFePO4 batteries, this will be most efficient mppt solar charger, not most budget sensitive, but definitely most efficient. I saw that Electra also designed mppt solar charger, and I'm interested to compare them both if we manage to get some more people to join us. Also we are also currently building is solar and battery powered sensors that can be used to detect forest fires. If your boards are available only to developers are there any plans to release schematics under some open source licence? We have experienced electronics people and we can at least try to build few on our own and share our experience and even help make it better. How many layers does your board use? Four or more? Cheers, Valent. On Thu, Nov 2, 2017 at 10:47 PM, Nicolas Pace <nico(a)libre.ws> wrote: