Introduction

Last week I was fortunate enough to get to properly test out the Daqri Smart Helmet at our offices rather than in the usual chaos of an expo.

Having had the Microsoft HoloLens for 10 months now and given talks on Mixed Reality I was most keen to try out the Helmet, see how well it performed and find out more about the roadmap for Daqri.

For those less familiar with the device, the Daqri Smart Helmet is a stereoscopic (binocular) Augmented Reality wireless headset built into a hard-hat. The device is focused on industrial uses and contains a large number of sensors to aid hands-free interactivity on-site in and around industrial machinery.

AR or MR

There is a somewhat of a grey area between stereoscopic Augmented Reality and Mixed Reality devices and both the Daqri and Meta2 (which this review does not cover) fall into it. Mixed reality is a term used to distance high-end devices that mix computer imagery into the physical world from phones, tablets, and eyewear that simply overlay (or augment) computer images in front of it. The first step to mix images into the physical world is to add depth by making images stereoscopic. Both the Smart Helmet and the Meta2 do this. The next step is to occlude virtual objects at a distance with real-world objects that are nearby. Both devices mentioned do not do this at the moment. In truth, there are no real fixed categories so don’t get too hung up on them.

Digital or Optical

There are two main types of Mixed Reality Headset:

  1. Those that combine the real world optically with virtual imagery using see-through screens.
  2. Those that use video cameras to capture the real world and then merge this with the virtual content to display on a regular head-mounted screen.

Optical devices offer perfect image quality of the physical world with no latency or low-frame rates as it is not being converted to pixels. This helps prevent any nausea for the wearer even if the digital content is not being rendered as fast as it should. A reasonably wide field of view of the physical world is much easier to achieve.

On the other hand, Digital devices offer the ability to see someone else’s mixed reality (from the headset) as the entire signal can just be switched. As both realities are displayed in the same way they blend together far better. Occluding digital content behind or interacting with real-world objects is much easier with digital MR. It is also a lot easier to switch to full VR by simply turning off the video feed of the physical world.

This review is centred around the Daqri Smart Helmet and other optical MR devices only.

Main considerations

While I don’t wish to imply that uses of the Smart Helmet completely cross over with the Microsoft HoloLens, it is the closest comparable device at this time, so making comparisons will help those that have already tried that.

Although I only got to try the Smart Helmet, I think it is important to note that this is the first form factor they have worked on, which works well for industrial applications. The Smart Glasses, which are also due for release to developers later this year, are likely to have a broader appeal and prove a closer competitor to other devices on the market.

 

 

 Daqri Smart Helmet vs Microsoft HoloLens Note about the up-coming
Daqri Smart Glasses
Appearance Similar in style to a US police motorcycle helmet, the device is highly visible and somewhat intimidating. For industrial usage, this is less of an issue. The HoloLens is somewhat less noticeable as it’s darker in colour and has no hard hat. Without the hard hat and front Visor the Glasses are much less noticeable than even the HoloLens.
Weight  It weighs 1.5Kg and although I was assured you could wear it for many hours, I was happy to be able to remove it after 15minutes. The HoloLens weighs just over 1/3 of this weight although does not have a built-in hard hat.  At 400g the Glasses are considerably lighter although do require a ‘pocket pack’ for power and other electronics.
Display The field of view is greater than the HoloLens although not enough to really strike you. The actual display delivered bright, colourful and sharp images although unlike the HoloLens I could see noise in the transparent (black areas). Likely to be the same as the Smart Helmet.
Tracking Tracking is what allows AR & MR devices to know where they are and keep track of your movements. Like the HoloLens the Smart Helmet employs inside-out ‘SLAM’ tracking to monitor your movements. Unlike the HoloLens the depth cameras are not employed for tracking only the wide-angle HD camera and accelerometers/gyroscopes. The better the tracking the more stable the virtual images are in front of you. Unfortunately, the virtual images in the Smart helmet were pretty jumpy compared the HoloLens unless you moved around slowly. I was told that the refresh rate was being increased to help with this. Likely to be the same as the Smart Helmet.
Features The Smart Helmet is designed for custom integration and usage and has a much larger array of sensors and ports than the HoloLens. Stand out features are the Thermal imaging camera, Barometer, Thermometer, 9-axis sensor fusion hub, beam forming mics and USB ports for external devices. Like the HoloLens it is also completely untethered with 64Gigs of memory. The glasses will not have the Thermal camera but most other sensors.
Batteries/Power The 5700mAh batteries are interchangeable unlike those of the HoloLens. During the 45minutes that the device was on, the whole device got pretty hot (on the outside) compared to the HoloLens that generates very little heat. I suspect the sensors and display draws a lot more power. Part of a pocket pack connected by a wire to the glasses.
Control/Interaction The Smart Helmet currently can only be controlled with gaze/dwell options and/or Bluetooth controllers. There is no hand gesture detection. New interaction devices are planned.  Likely to be the same as the Smart Helmet.
Price The Smart Helmet is currently priced at $15,000 for developers compared to $3,000 for the HoloLens. Estimated to be $5,000 and perhaps a closer competitor to the HoloLens or ODG Smartglasses.

Further information…

  • Daqri are working on being able to render graphics with visual-inertial algorithms which means even if it loses tracking it can carry on showing you what you are expecting to see.
  • The device still needs a starting point / zero reference using markers or manual positioning, depth maps cannot be pre-configured as with the HoloLens.
  • Main use case is in engineering, construction, heavy industry, plant management.
  • As with the HoloLens, it is possible to have someone, back in a control room, see what you see and offer guidance with voice and overlaid visual instructions.
  • Daqri are not working on collaborative synchronisation at the moment. This is a great feature of the HoloLens and useful in many business applications, I hope this will come.

Conclusion

Daqri have been working on an AR device for over 7 years now and the result is a very well-made and sophisticated product. To compete with Microsoft, ODG and possibly Magic Leap (at some point), they will need to keep iterating and focusing their uses cases. With just 300 staff compared to the thousands at Microsoft, differentiation and speed to market will be crucial.

The current cost really pushes it out of the hands of smaller companies or those not specifically needing a regulation/standards meeting device sealed to moisture, dust, shock, interference etc. Having said that, there are many multi-billion dollar industries including construction and engineering that may well see this as the only current device that can solve a lot of their problems.

While the hard-hat is a clever way to hold heavy batteries and extra sensors, it’s hard to imagine a significant proportion of Daqri customers choosing this over the lighter and slimmer Smart Glasses that are on their way.

Within a few years, Daqri are likely to be contending with a smaller HoloLens, ODG R9, Immy Neo, the Meta2 and possibly Magic Leap and Apple smart glasses, so now that most developers have emerged from their silos, its time for the real user-focused testing to begin.

2,825 total views, 0 views today

Share