Carman Neustaedter and Anthony Tang
CHI 2019 Telepresence Co-Chairs
CHI has created the opportunity for people to remotely attend the ACM CHI conference via telepresence technologies since 2016. The goal has always been to increase access to the conference for remote participants who would otherwise be unable to attend due to mobility impairments, chronic health issues, temporary travel limitations, or cost issues. This year:
- The primary way to remotely attend CHI is through the live streaming of talks;
- In addition, to support social interactions, we have created a programme that pairs up local and remote attendees through a mobile video conferencing setup.
Remote attendance has been carefully explored at CHI and other conferences. It has been carefully thought through and studied over a number of conferences. Ubicomp 2014 in Seattle, USA had 7 people attend remotely using Beam telepresence robots [1]; CSCW 2016 in San Francisco, USA had 19 people attend remotely using Beams [2]; and CHI 2016 in San Jose, USA saw 33 people remotely attend via Beams [3]. Each of these experiences was studied and, ultimately, it was found that remote attendees highly valued being able to use a telepresence robot to remotely be there, move around, watch talks, and engage socially with others. Beams worked especially well for small-scale social interactions with others, like during breaks and in-between sessions. Of course, the experience wasn’t without its challenges. Interactions were not always easy and sometimes local attendees were less excited about the Beams (e.g., blocking one’s view during a talk, being disruptive). But, for the most part, local attendees saw value in helping conferences create a more accessible and inclusive environment for people. Given the overall success, remote attendance continued at CHI 2017 and CHI 2018. Live streaming of talks was added at CHI 2018 in Montreal, Canada and this helped get around challenges with remote attendees finding it hard to see speakers and slides during presentations.
Telepresence robots present pragmatic challenges and other solutions are also needed. Over the years we also have faced pragmatic challenges with using telepresence robots at conferences. Telepresence robots require high Internet bandwidth over WiFi in order to work well. Beams (or equivalent telepresence robots) aren’t available everywhere and so we have had to ship them to each conference’s location from California, USA. This was not always cheap and did not present a globally responsive solution to reducing carbon footprints. While remote attendees helped promote sustainability by not traveling, as ironic as it is, the Beams still had to travel. CHI 2019 proved to be especially challenging when it came to telepresence robots for a number of reasons. Shipping costs to get Beams to Glasgow was high; the conference venue has a variety of levels, some narrow corridors, and different buildings making it hard for telepresence robots to move about; and, WiFi at the conference venue is not available at the bandwidth levels needed by Beams. (But don’t worry – WiFi should be fine for regular attendees! You likely don’t need the > 20 MBps upload and download speeds that a telepresence robot needs.) For these reasons, we decided to explore alternative ideas for remote attendance at CHI 2019.
Over the years, we have brainstormed a large number of different ways to support remote attendance at conferences, ranging from dedicated video conferencing tables in break areas and hallways to situated video conferencing links at the front of presentation rooms to programs that might pair up local and remote attendees. And, that is where we ended up going for CHI 2019.
Firstly, most paper sessions will be live-streamed, free of charge to people taking advantage of this. Secondly, we want to provide access to social interaction with our buddy/human proxy pairing of local and remote attendees.
Research has shown that ‘Human proxies’ can create an enjoyable experience for pairs of people. The idea started back in 2014 when two of Neustaedter’s grad students saw an episode of the US TV show, Arrested Development, where a man on house arrest uses a ‘surrogate’ to be present at work for him. The surrogate wore a head-mounted camera and could do things on the behalf of his boss. While meant to be humorous in the TV show, the students thought it would be fun to try the idea for real in one of their grad courses when one of the students was traveling and couldn’t be in class in person. The experience raised interested technical and social questions, so Neustaedter’s group decided to formally study the use of human proxies in more detail in two university classes. In the classes, pairs of friends tried out the human proxy experience; one stayed at home and attended class through a video conferencing link ‘worn’ by the other friend. The work was published at CHI 2016 and received an Honourable Mention [4]. One of the main conclusions from the study was that students really valued the experience of remote class attendance when paired with a friend. Classroom human proxies were not generally seen as a way to ‘control another person’; instead, it was seen as a way to do pairwise learning with a friend. Students liked being able to interact with their friend while engaging in the class and its activities. Of course, this type of experience still raises many interesting socio-technical questions and, done without the right intentions, could lead to challenging problems around ethics, privacy, autonomy, and much more. That is to say, we are fully aware of the complexities and issues with the idea of human proxies.
Creating networking opportunities through remote CHI buddies and proxies. When the challenges around remote attendance at CHI 2019 emerged, we thought that human proxies may be a possible solution to explore. We envisioned it as a networking opportunity where those local to the conference location could volunteer to ‘bring in’ a remote attendee using a wearable video conferencing system – e.g., a tablet that could be hung around one’s neck, carried in one’s hands, or held at one’s side. The local volunteer could take the remote attendee around the conference venue so they could interact with people and experience the social aspects of CHI like talking with others, seeing demos, etc. To see paper presentations, the remote attendee could watch the live streams of talks on the web. The benefit to the local volunteer would be that they could get to know more people at CHI by attending ‘with’ the remote person, and vice versa. Thus, we saw it as a valuable opportunity for both the local and remote person. We thought being a local volunteer could be especially valuable for newcomers to CHI such as new grad students who may not know a lot of people and want to network with others. It could also be a great way for a lab group attending the conference to involve one of their colleagues who can’t be there. Anybody who is interested can contact us and ask to participate as a proxy. There’s no intention of coercing anybody into being a proxy and proxies can certainly choose what they do to help out the remote attendee.
Given the number of people who have attended CHI remotely over the past two years, we anticipate that there might be up to about 5 human proxy pairs at any one time. Thus, it was not seen as something that would be implemented on a massive scale. Remote attendees will have to pay a small fee to help cover the costs of enhanced WiFi as well as the technical setup. It’s not a means to pay a proxy.
Overall, we see the use of ‘CHI local buddies’ or ‘human proxies’ as a positive bonus for both local and remote attendees to enable social contact and interaction. Our initial CHI web page describing the telepresence experience was unfortunately too ‘functional’ and did not clearly articulate our intentions or what we see as the benefits. The recent discussions have also caused us to rethink some aspects of the implementation by drawing in more volunteers and exploring different mounting/carrying solutions.
We are hopeful that our updates and this blog do a much better job of describing the intended experience and look forward to the community’s support in enabling increased access via remote participation at CHI.
References
[1] Carman Neustaedter, Gina Venolia, Jason Procyk, and Daniel Hawkins. 2016. To Beam or Not to Beam: A Study of Remote Telepresence Attendance at an Academic Conference. In Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work & Social Computing (CSCW ’16). ACM, New York, NY, USA, 418-431. DOI: https://doi.org/10.1145/2818048.2819922
[2] Carman Neustaedter, Samarth Singhal, Rui Pan, Yasamin Heshmat, Azadeh Forghani, and John Tang. 2018. From Being There to Watching: Shared and Dedicated Telepresence Robot Usage at Academic Conferences. ACM Trans. Comput.-Hum. Interact. 25, 6, Article 33 (December 2018), 39 pages. DOI: https://doi.org/10.1145/3243213
[3] Irene Rae and Carman Neustaedter. 2017. Robotic Telepresence at Scale. In Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems (CHI ’17). ACM, New York, NY, USA, 313-324. DOI: https://doi.org/10.1145/3025453.3025855
[4] Clarissa Ishak, Carman Neustaedter, Dan Hawkins, Jason Procyk, and Michael Massimi. 2016. Human Proxies for Remote University Classroom Attendance. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (CHI ’16). ACM, New York, NY, USA, 931-943. DOI: https://doi.org/10.1145/2858036.2858184
