Hi all,
I'm a newbie, who's keenly following all the activity in this arena. I haven't yet got my hands dirty with HW/SW but am just excited about the possibilities autonomous flight opens up.
I'm particularly interested in applications for police-forces and other security services, but think that there may need to be another stage in development in order to make the use of drones cost effective. At the moment, all the use of drones is line-of-sight, because of regulations. Eventually this will change as sense-and-avoid comes in, allowing drones to fly to the extend of their ranges. On top of that, the use of either 3G (or other) network means on-board radio comms/telemetry doesn't have to be direct to base.
The issue in terms of rolling out systems for law enforcement based on this, is that you still have the cost of the drone operator equalling the cost of a police officer on the ground. It's not scalable.
The only way to make the deployment of drones scalable is to have the both unmanned in terms of flight, and unmanned in terms of maintenance and re-charge. They need to have their own 'home' which allows them to re-charge without human intervention. Once you have this, combined with the other changes in regulations and network-based comms, a dedicated control-centre means a single user can deploy a dispersed collection of drones to an area much greater than if each drone had to be manned for each flight. The point is, you don't know where abouts your drone will be needed and when. My guess is that roofs of police stations would be the best place to locate each of the drone-ports, but you could also deploy drones straight off a drone-port on the roofs of cop-cars, meaning when a car is called out, you gain perfect surveillance whereever they go. Freeing up the cops on the ground to do their job, by running all drone activity from a control centre means you get the best of both worlds.
The key technology enabler for this is really accurate auto-landing, and recharging (and maybe tethering once landed, for any car-based drones). It seems that there's been some work done on optical landing systems, but I'm not sure whether anyone has been working on getting power back into battery packs by landing the drone onto charging pads, and having the recharging circuitry hooked up to the electrified legs of the drone. Obviously operation in wind is a biggy, but if you've got doors opening on the drone-port, this provides a level of help.
(As an aside, the other advantage to going control-centre based is that your operator can concentrate on flying the drone to the edge of its envelope. He wouldn't be standing on-site with a bad pair of wireless specs. He can be well-seated in a dark room with a decent joystick. This means better performance out of the drone)
My guess is you could have 1 pilot and 1 maintenance engineer cover 20-30 drones, which starts to make their use cost-effective. Just thought I'd throw this out there, to see if anyone has been working on the optical auto-landing for integration into the FW.
Thanks, Stuart
Replies
Check out skysense.. they are heading in that direction
Thank you, that's very interesting. I will get in contact with them. The first implementation of the autolanding system would (I think) just be a plug-in board between the receiver and the autopilot which can switch a channel 5 or 6 (or other) to say it has an ultrasound lock. This fakes taking manual control from a GPS autopilot in order to land on a pad automatically, so would in theory work with any existing autopilot off the bat. Cost of parts would be around $10-20 I think. Power consumption very low. Weight very low.
My pleasure. If you know anyone in the market for one let me know. I have discount codes for them. Let the know Eric from UAS Career Central sent you. Andrea (male) is the one to ask for. Eric
Good morning!
I am very interested in this type of autonomous UAV capability and I am new to the DIY "drone" forums. I have a concept of operations (CONOPS) that would incorporate this exact technology over thousands of miles of frequently inspected buried pipeline.
If you will, please keep me in mind as your research progresses. You can email me at davidaud@barrairpatrol.com. I will review your posts and more via this and other forums and will share anything I come up with as well.
Thanks very much!
First off, welcome to the community!
February of last year I started working a project that addresses exactly this: How to you make drones completely and utterly self-sufficient? We call these things autonomous vehicles, but they are only autonomous the moment the operator launches them. My dream is to make UAVs completely self-sufficient in that they will not need any physical assistance from any human operator (except for maintenance of broken parts), such that they launch, fly, land and recharge on their own. The only thing communicated to the vehicle is commands to launch or land or loiter over specific coordinates via a cell or satellite network.
In order for a UAV to have this capability, it needs a few important systems in addition to the flight-critical systems such as navigation, stabilization, control/communications, and propulsion. These systems are:
1) Energy Harvesting/Charging: Be it solar cells, charging contact pads, power-line induction devices, directed microwave or laser energy receivers, and so on: Whatever it takes to renew the energy supply in the UAV, which I presume is electric. If we are talking about gas-powered UAVs, then basically a fancy gas-station / drone-port receptacle.
2) Locomotion: Unless were are beaming energy to the UAV in flight or it is an eternal solar-powered vehicle, our vehicle must land to replenish it's power supply. Hence, it must either be incredibly adept at near-surface maneuvering, or have methods of locomoting its weight via surface contact, aka tracks, wheels, or legs. If our UAV is landing at a drone-port, then the design of the port must be such that the vehicle is captured and stabilized by the shape or systems of the port to enable charging or refueling. However, there is no need for a drone port if your plane is carrying solar cells: Although not efficient or large enough to sustain the aircraft in flight, the solar cells can allow the aircraft to "cook in the sunlight" while landed on the ground. Other options include hooking onto power-lines, using wind through the propeller as a windmill, etc. All of these will require some way of manipulating the landed position of the vehicle on its own prerogative.
3) Self-Launch: For VTOLs, this is easy; given there are no contact/interference hazards for the spinning blades or rotors, the vehicle just takes-off. For a Fixed-Wing, well, this is a bit more difficult, as I've discovered in my research in my project last year, but it is achievable.
The challenge is integrating these systems into a UAV design such that it is efficient and cost-effective. Initially, totally autonomous UAVs will be quadrotors / tricopters and other VTOLs, as these have the fewest problems to solve concerning launch and landing. They will have to be based to a charging station as you've predicted, since they don't have the range or efficiency to go far with much of a payload anyway. My dreams and designs center around an optimized fixed-wing VTOL Hybrid with solar-cells, for long-range missions that last from several days to months.
-Steve Carlson
Hi Steve,
Firstly, I'm sorry for not writing back when you replied to my mail. I've just re-read your mail and wanted to send you a PDF of a method of Ultrasound autolanding that I want to try.
Have you made any progress with you ideas?
Regards, Stuart
Quadcopter_Landing_System.pdf