Posted by Matt Johnston on January 14, 2010 at 7:44pm
Hi all,I'm a contracting WAN (large computer network) architect/engineer, and an embedded systems engineer (microprocessor-based stuff mainly, with most experience in netbook and mobile style applications). Before that I was a programmer, focusing on networked applications, and also did a while of AI development.I've been given a part-time funded project of the design of a mid-sized UAV (several to be built), for monitoring and on-demand surveillance of a medium sized property. Under the contract I have, I own all rights to the design, so I decided I'd share what I learn and develop with others.The planned UAV is a mid-sized quad-rotor helicopter, 500mm between rotor center and airframe center, with 200mm blades (for a 400mm rotor). These exact details may change however, and I'll try to keep this blog updated with progress and such.I'll be developing almost everything in-house, including the airframe, avionics, protocols, and ground station equipment. I'll be using pre-designed ICs, motors, and batteries (among other things) for obvious reasons. I'll be manufacturing a lot of the parts and assembly myself.The system is to be fully automated, and will be controlled via a base station server, with a small comms tower. The UAVs will contain mesh networking links, to allow either relayed communications with the base, or basic AI intelligence when they are outside of base range, by direct communication.I'll be posting to update on the sort of things I'm up to, and with more design and implementation details as they come up. I'll hope to follow this post soon with more information on how the system operates.Anyway, looks like a fun site, filled with a lot of knowledgeable people. I think I'm definitely going to enjoy my time here :)-Matt
Morli:
The main reason I'm reinventing is that the electronics would have to be all custom-made to support what I want (high processing power required, and advanced comms). I also want one designed as much to be able to handle bumping into things as it can anything else, since it's fully autonomous.
I do have some previous RC experience, so I know the basics, and I have a PPL (private pilots license), so I know the basics of flying (although I'm no heli pilot, sadly).
Xander:
Communications are going to be mainly done through a WiFi based mesh network (which if you're not familiar with, means that each node only has to be within range of one other, and they relay to each other, so you can have them far from base as long as there are nodes on the way back. Wikipedia has a good article).
Once the drones are given a task, they will be able to complete that task (including communicating with other drones about it) even outside range of the main base. It's intended that these things will/can operate with no humans in attendance.
Have any plans for how you'll be communicating with the quads? Given your background sounds like you'd be aiming for something fancier than R/C controls and analog video. Planning to get sufficient range to use wifi?
Matt, That looks like cool project to be in. All I can think of for now is some times it pays not to re-invent the wheel. There are few cool quad/multi/tri copter projects going on which might help you reduce the learning curve if you look around. Please keep us posted as I/we will all will learn a great deal from your experince
Gary:
These are primarily intended for mostly unpopulated areas, with the initial design (and the purpose it is initially contracted/funded for) being a farm. The device does have fail-safes, and will try to slow it's fall, and will have a siren output (although not too loud, the speaker is intended for other uses but would be recommissioned for this when needed.)
As for returning home, that's the idea. They measure how much drive battery power it took them to get from home to their target point, and after that amount of time plus a bit more is remaining, it takes them back home. I have a few ideas to make that work (electrically charged helipads, although a bit more complex than that).
Wow big project, how will you protect people on the ground in the event of an off?? Will they come home and charge themselves like those lawnmower robts??
Comments
What happened to your project?
The main reason I'm reinventing is that the electronics would have to be all custom-made to support what I want (high processing power required, and advanced comms). I also want one designed as much to be able to handle bumping into things as it can anything else, since it's fully autonomous.
I do have some previous RC experience, so I know the basics, and I have a PPL (private pilots license), so I know the basics of flying (although I'm no heli pilot, sadly).
Xander:
Communications are going to be mainly done through a WiFi based mesh network (which if you're not familiar with, means that each node only has to be within range of one other, and they relay to each other, so you can have them far from base as long as there are nodes on the way back. Wikipedia has a good article).
Once the drones are given a task, they will be able to complete that task (including communicating with other drones about it) even outside range of the main base. It's intended that these things will/can operate with no humans in attendance.
Thanks, I'll be sure to post updates often.
Gary:
These are primarily intended for mostly unpopulated areas, with the initial design (and the purpose it is initially contracted/funded for) being a farm. The device does have fail-safes, and will try to slow it's fall, and will have a siren output (although not too loud, the speaker is intended for other uses but would be recommissioned for this when needed.)
As for returning home, that's the idea. They measure how much drive battery power it took them to get from home to their target point, and after that amount of time plus a bit more is remaining, it takes them back home. I have a few ideas to make that work (electrically charged helipads, although a bit more complex than that).