Conference Musings

Written by Elecia White.

I recently attended a conference I wasn’t happy with. It got me to thinking about conferences, why I attend them, and what I want out of future ones.

I primarily attend conferences to learn things and to meet people. The things I want to learn fall into four groups:

1. Learn something in more depth than I can get from an online video or article (but not in so much depth as a book)
2. Acquire new ideas, one I’ve never seen before, such as new technology or upcoming academic research that will be important to my career in 2-5 years
3. Workshops where I can get my hands on some piece of hardware and get practical, actionable experience that I can use in my job immediately.
4. Infotainment (entertainment disguised as learning, or maybe learning disguised as entertainment, either way, I will learn non-critical things if the session is amusing enough).

Great sessions can combine all of these things (but they don’t have to).

There are also different kinds of people I want to meet: experts in their field (which often translates to potential guests for Embedded.fm), prospective clients or companies that might hire me in the future, past colleagues I only seem to see at conferences, friends, and listeners. I even occasionally find new friends but I admit I’m in “professional” mode not “social” mode for networking events.

I do enjoy attending the expo hall (and if I’m uncertain about paying to attend a conference or spending the time preparing a talk, I will often judge a conference on its expo hall). I like to talk to vendors, especially the ones who keep the same sales folk; I feel like I’m building relationships. I also use the expo to look for specific solutions to my current problems. While I tend to be dissatisfied with vendors giving session talks as though they were unbiased, I’m happy to occasionally sit and get a short session or a longer hands-on workshop from a vendor (see points 1 and 3).

I know a lot of expos have these “go to these vendors, listen to their pitches, get stickers/stamps, win something neat” conceits. I understand that, and though I do have a dev kit addiction that I don’t talk about, I do not participate in those. I’d rather engage with vendors I’m truly interested in instead of making the representatives give their spiel for practice. Given how many other people I see racing around to fill out the worksheets, I suspect I’m in a privileged minority (if I need it, I can buy it; if I ask nicely, I can often borrow kits).

I suppose this is a good time to say that these are (obviously) only my opinions. I am not planning to put together a conference myself, I’d rather program devices. I respect conference organizers and recognize it is a difficult job.  There are so many people with different opinions about what they want.  And I suspect taking all of my opinions would lead to an event that is not financially viable.

Embedded conferences are especially difficult because defining an embedded system depends very much on the person’s perspective: for 8-bit processor users, a BeagleBone Black is clearly not embedded; for computer users, BBB is certainly embedded. Even if that line in the sand can be drawn, we are still on the edge of hardware and software: is the conference going to cover both or be focused one way or the other? And then there are the industry differences— embedded devices go into many applications: the software approaches for medical, consumer, industrial, and scientific may be similar but they aren’t the same. For hardware, the approaches may be even more different.

There are different kinds of conferences. Tradeshows are a gathering of vendors and exhibitors; these can be helpful if you are comparison shopping for different processors, sensors, or other solutions. At the other end of the spectrum are the academic conferences and working conferences (e.g., IETF) where the goal is to meet with peers and exchange ideas. My ideal embedded systems related conference would be a combination of the two: a set of sessions and an expo or demo hall tradeshow.

I want hardware and software both. I’m not enough of an electrical engineer to follow deep dives into RF design (but I do want to be so I definitely don’t want an embedded software only event).

I suppose if I was in charge, I would organize tracks along application lines (robots and drone, medical, wearables) instead of technology or theory (RF design, low battery techniques, security). But that’s pedantry, what is really important?

I would like the talks to be practical, useful in the next six months. And while home projects are fine, I don’t want a big overlap with Maker Faire in the lecture portion (though I wouldn’t at all object to a mini-Maker Faire in the evening portion or the exhibit hall). Other interesting after-hours activities include five minute tech presentations (auto-transitioning slides are awesome) and bring-a-hack get-togethers that let people talk about their personal projects (and facilitate networking).

In the conference sessions, I want there to be a limit on advertisements. I’m happy to attend sponsored tracks but I want to know whether to expect an infomercial before I decide where to spend my time. And I have (and will continue) to attend vendor workshops, especially those with dev kits that I was going to buy anyway (or that I might use in the next 3-6 months).

At O’Reilly’s Solid conference, there were several short keynotes in the morning. The keynotes were generally more future looking and inspirational than technical. I liked that; it gave us attendees something to talk about at lunch (natural networking icebreakers). I’m not much one for inspirational talks (yeah, I gave one recently, I regret that). It is quite difficult to balance making a talk interesting and inspiring without making it a platitude fest. For keynotes, I’d rather hear about tech that is coming out in five years, it gives me ideas for directions I might take and subjects I might want to know more about.

It is tough to balance introductory material and advanced. Given where I am in my career, I want advanced embedded software and intermediate hardware topics. Too many intro sessions makes me feel out of place. Given the wide range of topics that fall under embedded systems and the need to have introductory, intermediate, and advanced topics in both hardware and software areas, the conference has to be either long, large, or focused on a particular subfield.

A long conference is exhausting. A large conference is difficult to put together. And a conference focused on a subfield might be poorly attended. Financially, putting on a conference is risky. How many people will attend and what will they want to see? How much are they willing to pay to attend? Is there some other event in the same time frame that will draw the same crowd? For conferences in their first few years, the answers to these questions mean the difference between success and bankruptcy.

Of course, the speakers are a huge part of successful conference. Many people want to evangelize for their product or company. Some speakers are there for the exposure, to sell their book or professional services (or advertise their podcast). Some want to give back to the community. And some are speaking because it is often a way to attend without paying the entrance fee. Even if the conference pays its speakers, putting together a new talk is a lot of work (easily 20+ hours for a 1 hour talk). In my experience, any travel stipend and honorarium do not compare to potential freelancing fees so there has to be another reason to attend.

Side note on one of my soapboxes: Organizers, please don’t complain about the lack of diversity, do something about it! Invite specific submissions and speakers, post the request for proposal in diversity specific groups and forums, ask people to suggest others (and invite them, I’m far more likely to submit a proposal or even attend if I know others there). Be proactive about diversity in your speakers, that leads to more diversity in your attendees.

Where was I? Ahh, when should you have your conference? Like most people, I tend to work for pay Monday through Friday and on my own projects on the weekend. If I attend a technical conference on the weekend, it will make me tired going into the work week. On the other hand, I would be more amenable to more intro, more hacker-y sessions on a weekend. How about a Sunday-Thursday conference where each day gets increasingly technical?

Given limited time and the stress of travel, I admit that I like electronic proceedings (videos!) that come out after some conferences. As a speaker, it is great to be able to show off a talk. However, I am much less likely to attend if I can sit at home (I see the need for networking and occasionally enjoy it but seldom look forward to it). Since the goal is usually to promote an industry and increase the knowledge (especially at academic or working conferences), videos make a lot of sense.

And that leads to my last point: conferences have to be self-sustaining (which means financially viable). An organizer has to balance the different types of income (attendance and exhibitor fees, usually) and the different types of spending (site, speakers, promotion, organizers’ salaries, etc.). There are four classes of attendees to consider: free (expo) attendees, paid attendees, speakers, and exhibitors. Each group has their own needs, some are conflicting. I’ve tried to represent the first three groups here but the exhibitors offset the price of the attendees so they are incredibly important as well.

Like a well-designed product, a conference with specific goals is far more likely to be successful than one with a more fragmented and indistinct approach.