ISMAR09: Workshop on Handheld Augmented Reality Games

This was a great workshop given by Blair MacIntyre from Georgia Tech on mobile augmented reality games. I got a lot out of it, from being reminded of some solid game design topics to getting new ideas about the game I want to make for my PhD.

The goal of augmented reality in this context is to embody social interaction in the physical world, enabled by a tight integration of the physical and virtual world. In terms of games, it's important to remember that design is more than just form and function - it needs context, too (which AR can give). Game design is about solving a problem within a set of constraints, and making something fun, challenging, awe inspiring, and captivating.

In augmented reality, mobility is usually assumed. But it's not just a combination of the physical and virtual world - there should also be registration between the virtual and physical worlds and real time interaction.

It's worth remembering that there are two classes of AR systems: task based and experiential. Task-based AR is perhaps not as well suited to handhelds, since your hands aren't totally free to complete the task, and it's hard to hold something light up for a long time. This is one of the areas that give head mounted displays (HMDs) an advantage: they can provide zero-effort, hands-free interaction and continuous peripheral information. Both interfaces provide some privacy, an in-place display, and per-user customization.

So what makes "good" AR anyway? What is unique about it that can be leveraged? Multiple people can work in a shared space, for one. Each person gets a unique view of the world while not giving up the global perspective. It allows for direct and natural interaction, and the physical world can be leveraged with props, spatial understanding, and dexterity.

Some of the graphics issues to consider when determining a platform for your game (cell phone or something more advanced?) include lighting, shadows, occlusions, and physics capabilities. Graphics don't always have to be real, either - non-photorealistic effects can help alleviate the processing power needed. Remember that latency is a bigger issue with AR.

Back to game design. We, as computer scientists, have to think like a game designer when coming up with new ideas. AR games shouldn't be all about the technology, turning them into demos, essentially. We need to create something that's fun to play. "The designer needs to envision how a game will work during play ... planning everything necessary to create a compelling player experience." In other words, you need to decide first what you want the player to experience, not what they will do, or learn, or whatever. This is a key point for me in thinking about my educational game.

The structure of a game includes the following components:

• players
• objectives
• rules
• resources (making the game not too easy, not too hard)
• boundaries
• outcome

Games are dynamic systems. They have a 'magic circle' defined by the boundaries. Dramatic elements make it emotionally engaging. Design is about balancing all these elements.

Remember that to make a game something more than a toy, there must be goals, and interesting and meaningful choices to reach those goals. The story and characters are brought out through actions.

Some questions to ask when making a handheld AR game:

• Who is your target player?
• When or where are they playing?
• Single or groups?
• Will there be props? How comfortable and easy to use are they?
• What exactly will the player do while playing the game?

Some of the constraints to consider when designing a handheld AR game:

• Fast motions are a problem.
• How will having the device (phone) in the player's hand affect things?
• It's tiring to hold up a relatively light device for long stretches of time.
• Awareness of other players.
• Small screens are tiring to look at for a long time.
• Vibrations and sounds to give feedback, especially when looking elsewhere.

The future of mobile AR gaming looks pretty exciting. We first need to make some good progress on tracking, and when we do, there will be some amazing opportunities to be immersed into a terra-scaled world. Plus, we will be able to create more social experiences and casual games, possibly hooking into social networks. How will you contribute to this future?