About AR Glasses
Xiaomi
2022
由于小米 AR 眼镜目前尚未上市,以下仅分享一些设计过程中的故事,均未包含设计结果等涉密信息
Since Xiaomi AR glasses are not yet on the market, the following only shares some stories in the design process, and does not contain confidential information such as design results
Research
如今主流智能眼镜包含以下三种形态
Today's mainstream smart glasses include the following three forms
Real-AR
1. SLAM测绘定位
2. 6DoF + 多组Camera
3. 厚重/分体式,不适合日常佩戴
e.g. Hololens 2
Pre-AR
1. 单目、单色 / 三色光机显示
2. 保留 Camera & 手势交互
3. 体积较小、重量较轻
e.g. Google Glasses
No Display
1. 无显示功能,触摸+语音交互
2. 音频+眼镜拍摄 + 轻运动场景
3. 造型轻便日常、时尚
e.g. Huawei Eyewear 2
每种形态间的关系如同商务本与游戏本一样,一方面是受众群体的不同定位,另一方面是对现有技术的无奈妥协;随着光波导与微型光机的技术进步,未来对于性能与外形的局限形态会随时间有所降低。
The relationship between modulation modes is similar to business games, and it is a different type of future promotion for the general public, which widely guides the goal of this and this technology wave; with the technological progress of light-mechanical low-light machines, there are more appearances for performance and appearance. will decrease over time.
我一直坚信对于一个优秀的产品而言,需要在特定时间点,通过稳定的技术,解决刚需问题,并稳固优化至最优解。在这个过程中,交互设计师是起到承上启下的关键作用。不光要对所有相关技术的底层能力有着明确认知,更要可以抽离出来找到用户的核心使用场景,探索更多可行性,从而追求最极致的用户体验。
I have always believed that for an excellent product, it is necessary to solve just-needed problems through stable technology at a certain point in time, and stabilize and optimize to the optimal solution. Interaction designers play a key role in this process. It is not only necessary to have a clear understanding of the underlying capabilities of all related technologies, but also to be able to extract the core usage scenarios of users, explore more feasibility, and pursue the ultimate user experience.
Exploration
Display
在传统界面设计中,我们通常只关注信息层级、颜色、大小、动态等二维平面构成。但在 AR 等三维空间中,我们还需关注空间性、安全性、以及易用性,并要在其中思考信息效率的最大化。比方说 AR 此时提示你前方有建筑物,但是你已经看到了,那么这个提示就是无用且具有干扰性的。
In traditional interface design, we usually only focus on two-dimensional plane composition such as information level, color, size, and dynamics. But in three-dimensional spaces such as AR, we also need to pay attention to space, security, and ease of use, and think about maximizing information efficiency. For example, if the AR prompts you that there is a building in front of you at this time, but you have already seen it, then this prompt is useless and intrusive.
Interaction
头显系统的交互相对于传统 APP 交互较为特殊,它像是一台电视,你无法直接触摸所对应的按钮或列表。但经过手势识别、动态捕捉等技术加持后,则可以通过你的手掌位置在虚拟空间中模拟出按压或拖拽的效果。
The interaction of the head-mounted display system is more special than the traditional APP interaction. It is like a TV, and you cannot directly touch the corresponding button or list. However, with the support of gesture recognition, dynamic capture and other technologies, you can simulate the effect of pressing or dragging in the virtual space through the position of your palm.
我使用 OpenCV 配合 Mediapipe 还原手掌关节点坐标实现的轻量化拖拽 Demo
使用 Unity Ar Foundation 框架 结合 Oculus Quest 2 进行的手势测试
Gesture
许多人认为未来的智能眼镜上的交互会偏向于拟真的手势交互,但经过多次测试后对于非专业场景的使用,我发现用户更偏向使用于小范围的按压或者触摸的交互行为(触控区通常在眼镜腿上)。他们的理由很简单,手在身体前动来动去又傻又累。
Many people think that the interaction on the future smart glasses will be biased towards realistic gesture interaction, but after many tests for non-professional scenarios, I found that users prefer to use small-scale pressing or touch interaction (touch The control area is usually on the temple). Their reason is simple, it is silly and tired to move the hands in front of the body.
电容式触控条被堆叠在极度狭窄的眼镜腿中,如何通过优化交互行为,在有限空间内使得操作符合直觉并高效易用尤为重要。
Challenge
AR 如何进行高保真原型测试?
How does AR perform high-fidelity prototype testing?
传统APP与网站的高保真测试我们几乎都可以在真机或模拟机上进行;但是对于AR眼镜来说,现在还处于初期阶段,对于硬件的局限性很大。在硬件方案没有完全确定前,设计师很难有趁手的工具进行模拟。对此我从形态上将其分为两种情况,Real-AR 原型测试 和 Pre-AR原型测试。
We can almost all perform high-fidelity tests on traditional APPs and websites on real machines or simulators; however, for AR glasses, it is still in the early stage, and the hardware is very limited. Before the hardware solution is completely determined, it is difficult for designers to have the tools at hand to simulate. In this regard, I morphologically divide it into two cases, Real-AR prototype test and Pre-AR prototype test.
Real-AR Demo
对于 Real-AR 的解决方案是采用 Hololens 2 + Unity 解决方案,借助 Hololens 2 的硬件能力与 Unity 成熟的 AR Foundation 工具进行开发测试原型。这套解决方案上手成本略高,但整体表现复合预期。
The solution for Real-AR is to use the Hololens 2 + Unity solution, using the hardware capabilities of Hololens 2 and Unity's mature AR Foundation tools to develop and test prototypes. The cost of starting this solution is slightly higher, but the overall performance is expected to be compounded.
Pre-AR Demo
对于 Pre-AR 就相对来说困难了一些,因为目前市面上难以找到合适的硬件设备可以支持呈现Demo。比如 Google Glasses 上市时间很长,性能有限不足以应对复杂交互与动效。其余设备要不就是单色光机无法现实彩色,外加不可运行三方程序,或者交互形式单一和交互与算力需要依靠手机进行。
It is relatively difficult for Pre-AR, because it is difficult to find suitable hardware devices on the market that can support the presentation of Demo. For example, Google Glasses has been on the market for a long time, and its performance is limited and not enough to deal with complex interactions and motion effects. The rest of the equipment is either a monochromatic optical machine that cannot achieve color, and cannot run a third-party program, or the interaction form is single and the interaction and computing power need to rely on mobile phones.
经过多方尝试后,我制作了一个专为 Pre-AR 高保真测试使用的程序,并通过 iPad 模拟 AR 实景的方案来解决这个问题。
After many attempts, I made a program specially used for Pre-AR high-fidelity testing, and solved this problem by simulating the real AR scene on the iPad.
这个方案的最大优点是可以通过 iPad + 摄像头模拟出用户使用AR眼镜的真实感觉,通过切换不同场景模版还可以测试方案对场景的兼容性。
The biggest advantage of this solution is that it can simulate the real feeling of users using AR glasses through the iPad + camera, and can also test the compatibility of the solution to the scene by switching between different scene templates.
对于交互方式来说,你可以根据需求,自定义选择使用手势、方向键、电容条等进行Demo模拟,极大的提升了原型制作效率。
For the interactive mode, you can customize the use of gestures, direction keys, capacitor bars, etc. for Demo simulation according to your needs, which greatly improves the efficiency of prototyping.