With Great Technology Comes Great Responsibility
RFID, or radio frequency identification, isn’t something most of us talk about in everyday conversation, but in many ways, this technology has become a regular part of our lives (though we often don’t see it). It’s in the books we check out of the library, the systems used to restock shelves at our grocery store, the passes that allow us to pay a toll on the highway without stopping, and the badges we use to swipe in and out of our offices.
In a world where maintaining the privacy and security of our personal information is becoming a complex challenge, the idea that others are collecting data about us is understandably concerning. That’s why I want to apologize for how AGU implemented and communicated a recent experiment with RFID at the 2014 Fall Meeting.
For those of you who are not already aware, here are the facts about AGU’s use of RFID at the Fall Meeting: The program was piloted in a small number of high traffic areas, including the exhibit and South poster halls, the Honors Ceremony, and in the large general session lecture hall (where named and union lectures occurred) of the Moscone Center. This was done to get an accurate count of attendance and to better gauge attendee participation. The sensors only counted people entering and exiting those spaces; they did not track an individual’s movements within a room, nor did they track movements outside of the designated areas within the Moscone Center.
While the data collected included name, institution, geographic location of the institution, and member type (such as student, associate member, regular member), no individual data will be used for any purpose or shared with any outside organization. In fact, AGU is only looking at aggregate data for all participants and breakdowns by broad demographics for these events. For example, we planned to compare the aggregate of students vs. regular registrants or international vs. domestic attendees, to help us understand the dynamics of large groups of people, as well as space needs and patterns of participation.
Screen shot of 2014 Fall Meeting registration webpage, including announcement about RFID program at the bottom of the page. Click the image to view full screen.
First and foremost, I want to apologize for the completely understandable privacy concerns our experimentation with this technology raised with some of our attendees. Clearly, the announcements we made during the registration process and on the Fall Meeting website about this program and how attendees could opt out were not sufficient. We could have and should have done better, and I can promise you that the lessons learned will be incorporated into our decision making processes – for all new technology implementations – moving forward.
Our initial decision to use RFID, as well as several other technology improvements, stemmed from a recent review that was done for AGU’s entire meetings program (something we do regularly, not just with meetings, but with all AGU programs). Having good data can significantly improve meeting planning, and many organizations that present large conferences use RFID technology to assist with their efforts. For AGU, having accurate attendance counts and understanding the types of members attending our events can help us to better plan for things like room sizes and refreshments. Getting accurate counts has been challenging for us in the past because it has been done through the use of manual clickers or visual counts, which is not efficient or effective.
Because we heard concerns from some meeting attendees, I want to let you know that individuals who are uncomfortable having their information included in the data set can request that it be removed. Just send an email to [email protected] and we will delete your record. Also, I want you to know that we recognize that member feedback should inform our decisions, so if you have any further thoughts on how we should could or should use this technology (and others), or how we should communicate about these issues, I welcome them. We will host a question and answer session for AGU members via conference call on Thursday, 29 January at 2 p.m. EST. If you are interested in participating in this session, we ask that you register by sending an email to [email protected] – only so that we can send you information on how to participate and ensure that the technology will support the number of participants.
Of course, if you are unable to attend that session, you are welcome to send your thoughts to us via an email to [email protected] or through the comments section below.
Please know that we regret our poor execution and communication of the RFID experiment and we look forward to working with you to develop a path forward with which we can all be comfortable.
Very nicely done piece on the badges. Just the right mix of being apologetic, letting folks know exactly what was done and why and asking them – hey any ideas to help us do it better – we are listening.
The first rule is: avoid unnecessary data. If the RFIDs are used to determine how many people enter a room, then it is absolutely unnecessary to collect name, institution, geographic location of institution and member type.
I tried to get my data deleted as explained above and on 16 Jan I got the response:
Its 23 Jan and still no word from the Meetings Department. I realize deleting my data is mostly symbolic as it has probably already been logged and distributed but AGU should follow through with its offer in a timely and expedient manner.
I apologize for the confusion. Your email came in before we had set up an internal protocol for managing these requests, which is why you were told to contact the meetings department. We are removing your information from the data file. All requests sent to [email protected] will be dealt with directly.
You will receive a formal confirmation email next week, but if you have any additional questions in the meantime, please let me know.
Manager, Strategic Communications
I wonder how accurate the RFID data on say, for example, attendance at a named lecture will be? For example, I was delayed getting to the Stephen Schneider Lecture, but after stepping into the room and seeing how full it already was — and the difficulty I would have had climbing over people to get to an interest seat — I decided to leave the room with plans to watch the lecture at a future date through the Virtual Options website. So, is my entry into the room — even for just a minute or so — counted by the RFID logger as an “attendee”? How many other folks might have been in my same situation? Why not just have an affirmed count of the number of chairs in the large lecture halls prior to the event, then “eyeball” the % of seats filled at various times during the event. I would think this would be more than adequate to inform decisions on whether a bigger (or smaller) room is necessary for next year’s event.
I am a US citizen and I wonder what the meaning is of the word “ALIEN” which is printed on the RFID sticker, on the same hidden surface where the chip is attached. It is visible if you open the badge and place it against a source of light. I asked my colleague, also AGU attendee, and his RFID sticker has this word too.
Just wanted to let you know that ALIEN is actually the name of the manufacturer of the RFID chip that was used: http://www.alientechnology.com/.
Manager, Strategic Communications
OK, so this is the name of the company that benefited from this absurd project. Good to know. Now imagine what our foreign guests were thinking, those who noticed this designation on their badges.
Being tracked and monitored has become as American as apple pie – unfortunately. It makes a LOT of people uncomfortable especially the amount and type of data collected. Just because a technology might exist is not sufficient reason to use it. Checking out a book from the library is a very different activity from personal scientific discourse with new and old colleagues or tracking the proximity of one identifiable person to another. If you want to count people, give the badge monitors clickers.
The information about the program was insufficient. I certainly either did not read or see the information about it which tempered my outrage when I saw the outline of the antenna in the badge. I perhaps should have but it was not only myself who was unaware. Absolutely everyone I mentioned this to was surprised and immediately removed the stickers and left them in various places.
This should never be used at AGU meetings.
I had not seen either the notice in the registration page, or the RFID sticker on my badge.
So here, after the fact, I found out about this. And with the now-usual admonition that I have to “opt-out” of something that I never chose to do. Pretty invasive, and pretty silly. And what was the overriding reason? Getting a count of people at sessions? Huh? You claim it is inaccurate to simply observe the selected rooms? How is it Inaccurate to judge whether a room is crowded or not, just by observing? And you overcollected data in the now-common US tradition. Collect more data than you genuinely need, then forget about any privacy concerns once you’ve collected it.
This was a badly designed and badly carried out experiment.
I am frustrated that you did this.
Again, my main question is WHY?
Thanks for this post.
In revisiting plans for next year, I would suggest that the technology investment that would most improve the AGU experience for members and attendees is not RFID, but would rather be a simple web interface and app that actually work to quickly identify the talks and sessions of interest.
There has been incremental improvement over the years, but it is still a slow, inefficient, and painful process.
I spend at least 3 hours each year manually searching the same names and topics. It is difficult to get a quick overview of the whole meeting (e.g., session names appear only by day rather than overall by section, so a lot of clicks are required to see them all.)
Why isn’t there a way to save my research topics and researchers to my profile from year to year, to easily share my profile with friends with similar interests, and to have suggested talks highlighted that match my listed interests? This would save a lot of time.
If each of the 24,000 attendees is spending 3 hours battling the online program, this is 8.2 years of collective time that we are not spending on science, teaching, or communication. Surely it is well worth a few weeks of a few programmers’ time to make improvements- and would be well worth testing with a user group to make sure it meets members needs.
It is illegal collect such data, and inform participants only after the meeting. Also, I do not see reason why organizers used registration fee for this.