Académique Documents
Professionnel Documents
Culture Documents
Robots are all around us. For years weve seen them in industrial applications, deployed
as precision factory workers that need no lunch break. More recently weve seen them in military
applications for reconnaissance and for entering places too dangerous for humans. Most recently
weve seen robots at work in on behalf of public safety agencies, defusing bombs after the
Chelsea, New York bombing and as a delivery mechanism for a grenade that ended the life of the
Dallas sniper.
Why not put robots to work in journalism? There has been much fascination with
unmanned aerial systems (Ill use this term interchangeably with robot) as a replacement for
expensive helicopters, but less so with ground-based systems, which can be equally effective in
their own way. Unmanned ground-based systems can go to places where human journalists cant,
shouldnt or wont and can offer a unique visual perspective.
Professional-grade robots cost between $10,000 $50,000 and up. There is some
programming required to customize them.
As is often the case with new high-tech hardware, the immediate assumption is that only
large media companies will have the opportunity to leverage robots in the near term, mainly
because of the assumed high cost and technical expertise required to deploy and operate an
unmanned ground system. For big stories covered by national and international news in
dangerous places like war zones, disaster areas, public demonstrations and active crime scenes,
there are many uses for robots. For regional news, there might not be as many of these types of
stories.
Hypothesis
For my field test, I asked the question: Can robots be a feasible option for cash-strapped
regional news organizations?
There are several important factors in answering this question: 1) Is there a cost-effective
way to build and control a robotic platform? 2) Can someone with a moderate level of technology
savvy build and operate that platform? 3) Can that platform easily support affordable, easy-to-use
existing technologies? 4) What types of stories can be supported by robo-journalists?
Regional newspapers audiences are interested in localized stories that offer a unique view
into the places they see every day. Robots could be one tool to provide those views.
Separated into these smaller projects, the task became easier. Assembling the entire
device took 2 hours and 48 minutes. Someone with slightly better technology skills likely could
have cut that time in half.
The Camera. The smart phone mount was designed with a sticky padding to which the
phone could be affixed. I quickly learned it would not be enough to keep the phone stable while
the robot was moving and that it was designed to mount the phone vertically, which is not idea
for capturing photos or video. I solved this issue with a rubber band and some Velcro. Since the
camera is only a small part of the phones surface, I was able to tightly secure the phone to the
mount in either orientation with a rubber band around the belly of the phone and small pieces of
Velcro at the corners to keep the phone straight. Mounting a smart phone to the robot was
important as it is by far the most familiar piece of technology for a journalist at any technical
level. This also opened up the ability to add functionality to the camera via the scores of
accessories and thousands of apps that could be added to the phone, rather than being dependent
upon soldering or attaching new components to the Arduino board.
The Light. If the robot was to enter small spaces or be used at night, the iPhone
flashlight would not provide enough light. There are light kits available to install on the Arduino
board, but keeping to the rule of making this kit easy, I went back to the Velcro method: attaching
an LED light designed for an SLR camera just in front of the smart phone. The switch would
have to be manually turned on before sending the robot away to do its work.
The final product looked like this: http://bit.ly/2cwqxiK .
Each had its weaknesses. The USB cable was the best way to program the robot, but not
practical in the field. After downloading and trying many different iPhone apps, I learned that the
Bluetooth Bee is not compatible with the iPhones Bluetooth. This left me with the IR receiver as
the only practical option. The challenge with the IR receiver is that I would need to either find or
write a Sketch (the programming language for the Arduino board) in order to make it work.
I found a simple SparkFun IR remote back at RobotShop.com for $4.95 (http://bit.ly/2dwoVKN ).
Creating the Sketch to make the IR operate the robot was a much more challenging task.
After searching the RobotShop app store (http://bit.ly/2cX21dx ), I discovered there was not a
turn-key app to make the robot go forward, backward, right and left; to control speed and to
actuate the servo motor that tilts the camera. The beauty of open source technology (Arduino is
open source) is that there is a community of people who love to solve problems. I learned that for
open source projects like this one, the Android operating system is preferable to iOS as there are
more Arduino apps available on Android, and more turnkey Sketches written for Android-based
apps. I posted my issue on several Arduino discussion groups and finally, someone sent me a
template for creating a Sketch. After a quick study and some Sketch tutorials, I figured out where
the blanks were in the code (the aforementioned statement summarizes hours of study). It turns
out I just needed to enter which pin the IR was plugged into and which pins each motor was
plugged into, then assign commands to the parts of the code that described the buttons on the
remote. Once I figured that out, it took 5 minutes to customize the code, though I did not figure
out how to control the speed of each motor. Each motor was either on or off full speed or no
speed.
Remote Operation of the Camera. Given how important the iPhone was to the project, I
needed to find a way to control the iPhone mounted on the robot with another iPhone held by the
journalist or robot operator. I tested the combination of BluxPro/BluxLens
(http://bluxtouch.com/camera/ ; $0.99 in the app store), which turns one iPhone into the camera
and the other, remotely into the lens for the camera. The two phones were paired using WiFi. I
quickly found two weaknesses with the app 1) Both phones had to be connected to the same WiFi
network, a problem outdoors, even if one iPhone had its personal hotspot function on, which was
a slow and unstable connection; 2) The Blux product is excellent for still photos, but did not
allow for remotely capturing video. I had better luck with CameraPlus (http://apple.co/1WicWk6
; $0.99 in the app store), which has developed an Air Snap connection that utilizes WiFi and
Bluetooth to pair the phones. The connection was still a bit slow, but was stable. The app also
allowed the remote operator to switch between video and still photos.
feet away. It was a crabs eye view of the pre-hurricane sea. Or for city dwellers like me, a Rats
Eye View. The robot now had a name The Rat. It was a scruffy little beast that went places I
didnt want to go and saw things from a low-to-the-ground view.
Lets say youre a local paper and youve decided to do a story on litter on the towns
street sanitation. You could take pictures of garbage bags piled on the street, maybe show some of
the little people leave and the residue left after garbage day. From a photographers view it would
be an interesting story. From a rats eye view, it becomes difference. Its a macro-lens view thats
hard for a human to effectively catch. Its not a world-changing story, but its a good test for The
Rat. So I went out on a Thursday night garbage night in my neighborhood and had The Rat
offer a Rats Eye View of Garbage Day in New York. It appears here: http://bit.ly/2dbc0Mf . The
Rats Eye View offered a slightly gross, but impactful perspective.
Back at Washington Square Park on a beautiful day, I took another picture, this time in at
a quieter, more controlled time. The result was beautiful. If the iPhone had had a 360-degree
snap-on attachment, it could have captured a nice scene in New York life. I would not have had to
worry about where to hide myself to stay out of the picture as I could run The Rat far enough to
allow me to duck behind a tree, and the low-to-the-ground position of the camera would have
captured a more complete scene than is possible from a higher stance http://bit.ly/2du2y61 .
My last test was one of sending The Rat somewhere I couldnt go. I found a huge
construction site almost a half-block, surrounded by a wooden privacy wall with a cloudy
Plexiglas window to peek in. I searched the perimeter and found what looked like a square cut
into the wooden wall to allow cables to be fed into the site. I ran The Rat through that doghouse
about six feet into the site and snapped a picture. The Rat got stuck for a minute when the
concrete ended in a patch of sandy, gravelly earth, but eventually, I got it back out (Ill admit to
having tied a piece of twine to The Rat as a just in case measure. Heres the difference in what
we saw: http://bit.ly/2dwDrCo .
I did not test the idea of using IFTTT or a similar app to automatically forward video and
photos captured to an editor, blog or social network, but that is a simple thing to try, should
someone wish to do so.
1)
Motor noise. The small motors on the robot emit a high-pitched sound making
it difficult to capture video with sound when the robot is moving.
2)
Tracked wheels. When running the robot, the tracked wheels made for a very
bumpy ride, making video and images captured unusable. Next time, I would
use sturdier (this one had a lot of plastic) chassis with large wheels, each with
its own motor. This would present a programming challenge, but would absorb
shock better and allow for better maneuverability.
3)
People freak out when they see something mechanical running in public. I
didnt think to put a smilie face or a bicycle flag or something else to make the
rate cute when I used it the first time in Washington Square Park. It was
jarring for people to look down and see a mechanical item with wires hanging
out and a phone attached in their midst. Its important to make visually clear
that this device is not intended to be sneaky or sinister. A waist-high flag
attached to the unit and some sort of cute sticker or other dressing would help.
4)
iPhones are not great for do-it-yourself projects. As mentioned, there were
many apps for controlling an Arduino board, some for iOS, but most for
Android. The Bluetooth Bee was also not compatible with the iPhone, so the
separate IR remote was needed.
5)
Two remotes is a lot to juggle. With one remote to operate the device and the
iPhone to operate the camera, it was a bit unwieldy at times.
Further Reactions
I introduced The Rat to my Facebook audience to test reactions. One I hadnt thought of
was a comment that I would cover the story of the robot getting arrested for upskirt
photography. This is a valid concern. Use of small robots like this would need to be done in a
way where people know the robot is in use, perhaps by incorporating a light to indicate if the
camera is on. Either way, the operator must be careful to avoid invading peoples privacy.
Another privacy concern came from a commenter who suggested using The Rat to
capture private conversations about politics in restaurants to capture the real thoughts of voters.
This would require telling diners that The Rat is there and ensuring there is explicit permission to
use it for such a thing in order to avoid legal issues.
Other ideas generated mostly focused on peoples fascination with seeing places they
cant go ranging from the sewer and other hidden city infrastructure to behind police tape lines.
Conclusions
Based on my initial hypothesis, following is my assessment of the test:
I believe that unmanned ground-based systems are an opportunity for regional news organizations
to apply innovative newsgathering technology either before or on the same timeline as larger
news organizations. It can be inexpensive, easy and unlike unmanned aerial systems, these
devices do not require an operators license.
Next steps
Continuing this project falls into three main areas:
1) Perfecting the robotic platform. The next versions of this process should focus on
reducing motor noise, bumpiness when the robot is in movement as priorities. There
10
sensor and sound detector on a robot and when theres a boom, shake or fire it can go find the
source and film it on its own.
Imagine crime reporting where instead of delivering lethal devices to a criminal, a robot
can enter a crime scene, capture high-quality photos, broadcast them and crowd source the perps
identity, even before an arrest can be safely made.
Were at the beginning of using both aerial and ground-based systems to gathering and
understand news. There is much more to come as these technologies are put to use.
11