[Osmf-talk] Notes & Bing Imagery

John Crowley bostoncello at yahoo.com
Tue Dec 3 17:16:07 UTC 2013

FWIW, for the past four years, I have curated quarterly field experiments with a wide range of UAVs: over 26 airframes so far, ranging from small handheld quadcopters to 4m-wingspan aircraft and even a Cessna 337 that has been rigged with Predator A optical package. 

Yes, the hope for UAVs to provide imagery is real. They offer a wide range of capabilities, with EO cameras with 3cm resolution or better. Some have capacity for capturing infrared or multispectral imagery. Some beam all this data down in real time.


Let me set aside the privacy issues, not because they are unimportant, but because we are making good progress in learning about them. I am more worried about something that gets missed in these discussions: UAVs raise myriad public safety issues.

1. They crash. A lot. 
Even the ones that have pre-programmed flight paths (there is a story of a well-known small airframe crashing into one of the largest informal settlements in Haiti, and I personnally watched a $50K carbon fiber "bird" crash into a fence post on approach to land). Changing weather conditions, interference with flight controls from unexpected RF emissions on the same channel, and even loss of comms with the bird all happen in our experiments. 

2. UAVs lack see and avoid capabilities. 
Expecting these small birds to operate in areas traditionally used for manned airspace is a hazard to private pilots and commercial and military flights. Urban areas raise risk of crashes that would injure citizens. One *must* control the 3D airspace and provide constant updates on the position of any UAV and its envelope in that airspace to everyone. Most private pilots lack even basic tools for such tracking, and must instead rely on radio contact with someone who does. 

The procedures for reducing these risks are going to take a while to work out. Look up issues on temporary restricted airspace and the expense that the US government carries to scramble fighter jets when private pilots wander into these temporary zones. This is not a technical challenge, but one that involves a huge community of private pilots whose airframes and communications packages may be decades old and ill equipped to handle the demands of mixed manned/unmanned airspace with significant (costly) upgrades (costs which might have to mandated by governments as regulations--an imposition many dislike from the start).

3. Municipalities lack regulation on UAVs.
Many cities and subnational are beginning to experiment with UAVs. They don't yet have policies around how to handle UAVs that crash into people, buildings, or other aircraft. They will not be likely to take a "try it and we'll see how it works out" approach for OSM. Flying a UAV over London will probably not yield a positive municipal reaction. Flying one in Moscow... not a good idea. Or someplace where even the accusation of foreign espionage can lead to jail. It is an activity that must be done within local rules, laws, and even just political support.

And then there is the technical problem of mosaicking hundreds of thousands of these still photos into something that remote mappers like the OSM community can use. It's a solvable challenge, but difficult and often not fast enough for response work.

We should be excited, but we also need to realize that UAVs are not the same as code. Refactoring after a mistake involves fixing not just broken technology, but potentially broken (or jailed) humans as well. They are changing fast and growing in capability, but they are still tools to be used with great care around safety of flight.

John Crowley

On Tuesday, December 3, 2013 10:43 AM, Oleksiy Muzalyev <oleksiy.muzalyev at bluewin.ch> wrote:
I just received a reply from SenseFly:

-- QUOTE --
The eBee is priced CHF 19’990.- + shipping/taxes. This includes
everything needed to fly (ebee, camera, accessories, batteries,
chargers, spare parts, software, hard-shell case.) except the laptop.
The package also includes the picture processing software in order to do
orthomosaic and DSM.

Please note that our products are tools for professional and none RC

I can tell for sure that pilots must be formally trained in a company or
an organization to fly such an expensive aircraft.

And there should be several pilots for one drone, so that it is
constantly in the air, moving from one city to another. Perhaps one
drone for a country.

O.M. (Alex-7)

On 03.12.2013 16:27, Clifford Snow wrote:
> On Tue, Dec 3, 2013 at 12:16 AM, Oleksiy Muzalyev
> <oleksiy.muzalyev at bluewin.ch <mailto:oleksiy.muzalyev at bluewin.ch>> wrote:

> I believe their web site mentioned a price of $12,000 US. It does make
> mapping some areas much easier than taking gps readings. Building
> outlines, roads in gated communities, rivers and streams hard to reach
> locations are just some of the thoughts I had about using a drone. What
> sets SenseFly apart from others is the automated flight capabilities. As
> a talk at this years SOTM-US shows, you can put together a drone much
> cheaper that can produce images that can be orthorectified.
> In the US, there is uncertainty, Jeff Bezos aside, of the legality of
> flying a drone. Rules are due by 2015. 
> -- 
> Clifford
> OpenStreetMap: Maps with a human touch

osmf-talk mailing list
osmf-talk at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/osmf-talk/attachments/20131203/d7acc9ca/attachment.html>

More information about the osmf-talk mailing list