Blog

categories

tags

Three Takeaways on Sensors at the Online News Association Conference

It happened! Senselab built 30 sensor boards and placed them around the keynote room at the Online News Association’s Conference. The data they collected flowed up through dropbox to a visualization developed by the Data News team at WNYC.

senselab

Kipp Bradford and Julie Steele of SenseLab, showing off the finished sensor board for the ONA13 conference, with motion detection, temperature, air quality, cell phone radio frequency, and noise level.

 

 

SensorViz

The visualization of sensor data that WNYC’s data news team built for the ONA13 Conference

What did we see? We haven’t done the full data analysis yet, but it was easy to see pockets of high-cell phone data use (as per the image above; taken during our panel),and that during the surveillance keynote the sensors on the right hand side of the room were reading lower temperatures than the left-hand side.

In the next week we’ll be publishing the code, the board design, the list of parts and a more complete walk-through of the process, but in our Saturday afternoon panel, we heard a lot of demand for resources for people who want to understand sensors and their possibilities.

It’s a really fast moving field; but there is help available.

Here are some resources to learn the tech:

Kipp Bradford, the engineer who worked on this project, remarked that the steps and effort that required 10 people working for multiple weeks last year took only one person working for a week this year. That’s how quickly the resources and technology are developing; a great leg up for people just starting out. You can move fast by building on tools already produced by the community.

There are plenty of resources for people starting to design and build physical computing with sensors.

Learn Arduino – Adafruit, who sell kit for makers, have written good tutorials to learn Arduino; the most common electronics prototyping platform. They start at ‘Lesson 0: Setting Up’ and the latest online tutorial, number 17, will teach you how to build a motion sensor which triggers an outbound email. Using Arduino is a pretty good way to get your head around electronics and start to understand the possibilities. Keep in mind that other platforms for physical computing do exist, with different strengths and weaknesses.

Reach out to a maker community – These are the people who can help you on your journey; whether it’s navigating problems, or generating ideas. Maker and hacker spaces exist all around the world; North America hosts more than 250 alone. If you’d like to read about the culture of hackers before you dive in, try this: Coding Freedom

Hackers have put together spaces from Baghdad to Baltimore

Hackers have put together spaces from Baghdad to Baltimore

Read Make Magazine – This is the maker community’s magazine of choice – everything from knitting and craft to 3d printing and woodworking. They cover the physical computing world with gusto.

Get deep into sensing – Kipp referred to Sensors — Open Access Journal: It’s a pretty advanced read, aimed at a professional, specialized readership.

Here are some resources on the intellectual side of sensing:

The links above cover off practical and technical resources, but we’re keen to focus on the intellectual tools as well. The Tow Center is in the process of writing a report covering the context and landscape of sensor journalism. We’re also commissioning research on the legal and ethical issues involved. Meanwhile, here are some of the readings we’re reviewing.

Learn good experiment design – The EPA’s Guide to Quality Assurance Plans is aimed at volunteers who monitor local water sources, but their basic quality control concepts: precision, accuracy, representativeness, completeness, comparability and detection ranges, are relevant beyond pollution sensing.

Delve into epistemology; the study of truth and knowledge – This wonderful article called What Does the Zapruder Film Really Tell Us? about film footage of the Kennedy assassination should whet your appetite.

Different Projects Have Different Requirements for Accuracy.

The sensors we deployed in the keynote room could give us a “good-enough” understanding of the variation within the room; we knew the relative warmth of some areas, we could collect data on motion, and we could see areas of high cell-phone activity. However the goal of our project was not to produce absolute measurements that could be compared to legal limits on environmental conditions or known health effects.

By comparison, USA Today’s Ghost Factories project concluded that the contamination left behind by specific closed factories posed specific health risks. That relied on precise, absolute and verifiable measurements, produced with rented $41,000 kits and cross-checked by Howard Mielke at Tulane University.

The WNYC Cicada Tracker, despite involving uncalibrated sensing, mobilized a strong community of people who learnt about how Cicadas interact with their environment, learnt some technical skills and became sources for UGC data. The undoubted success of the project didn’t rely on evidence-grade equipment, but it did rely on engaging listeners and providing an accessible way to participate.

 

Sensing is Linked Closely To Surveillance and Privacy Issues

Circuit boards with blinking lights under seats were conversation starters. Journalists’ curiosity prompted them to read the short explanation sheet we attached to each one and – as far as we know – no-one was too bothered by it (after all, the measurements we were taking are fairly benign), but we overheard plenty of jokes and saw the odd tweet about surveillance.

We deliberately invited the discussion; one of the sensors we set up was collecting the MAC addresses of every switched on WiFi device in range. Why?

There are already businesses using MAC addresses as a sensed proxy for tracking individuals’ presence. Here’s a paper saying how it’s done (pdf). Cameras (a sensor type we’re all familiar with) backed by facial recognition software, can achieve the same effect, although both techniques risk false positives and negatives.

Journalists collect a lot of personally identifiable information, sometimes including the most intimate details of a subject’s life. However, as we consider adopting technology which can harvest larger amounts of data – sometimes indiscriminately – we’ll need to consider the implications and responsibilities of collecting and storing it.

Here are some readings we’re using to get our heads around it.

Sousveillance –  Inventing and Using Wearable Computing Devices for Data Collection in  Surveillance Environments, by Steve Mann, Jason Nolan and Barry Wellman.
“This paper describes using wearable computing devices to perform “sousveillance” (inverse surveillance) as a counter to organizational surveillance. A variety of wearable computing devices generated different kinds of responses, and allowed for the collection of data in different situations. Visible sousveillance often evoked counter-performances by front-line surveillance workers. The juxtaposition of sousveillance with surveillance generates new kinds of information in a social surveillance situation. ”

If Personal Information is Privacy’s Gatekeeper, then Risk of Harm is the Key: A Proposed Method for Determining What Counts as Personal Information‘ by Eloïse Gratton
Suggesting data protection laws are written using ‘personal information’ as the key concept, and proposing that ‘risk of harm’ is a more relevant measure.

The PII Problem: Privacy and a New Concept of Personally Identifiable Information by Paul M. Schwartz & Daniel J. Solove
Noting that information that was previously considered unidentifiable can now be de-anonymized, rendering the concept of Personally Identifiable Information more of a continuum that a binary measure.

Fergus Pitt is a Tow Fellow working on the Tow Center’s Sensor Newsroom Project at the Tow Center for Digital Journalism.  The Sensor Newsroom Project is a project made possible by generous funding from both The Tow Foundation and the John S. and James L. Knight Foundation. The goal of the Sensor Newsroom Project is to explore how recent advances in sensor networks, citizen science, unmanned vehicles and community-based data collection can be used by a new generation of sensor journalist to move from data analysis to data collection. Follow Fergus Pitt on Twitter @Fergle. To learn more about the Tow Center Fellowship Program, please contact the Tow Center’s Research Director Taylor Owen: taylor.owen@columbia.edu.