DAQRI, a leading AR (augmented reality) innovator, recently introduced its Smart Helmet product, billed as: “The World’s First Wearable Human Machine Interface.”
We spoke with Matt Kammerait, Vice President – Innovation, of DAQRI, about the Helmet, the company’s strategy and the future of AR and related areas, including robotics.
Smart Helmet – What’s So Special
The Helmet has been designed to fit into a wide range of industrial settings, where workers have to interact with machines and an often difficult environment. It has a hard hat exterior; a face shield; an array of cameras, including infrared and HD; four microphones; and a sensor package that enables 360º recognition of the environment around the worker.
The Helmet allows the worker to view the actual scene before them, while the AR also displays information about the items they are viewing. They can communicate and call up relevant information or assistance as needed for their task.
Kammerait describes the Helmet as “well-balanced, very protected and designed to slot into the user’s existing ecosystem.” This is critical, because in typical use cases the Helmet wearer will be drawing information from the existing system of the company or facility, as well as providing new input.
The objective was to make the Helmet usable and safe, even in extreme environments and to make operation as hands-free as possible.
Inside The Smart Helmet & Its Apps
The Helmet uses an Intel RealSense camera package, which Intel claims can, “‘see’ like the human eye to sense depth and track human motion.” It also incorporates Intel’s Sixth Gen Core M7 processor. DAQRI’s trademarked Intellitrack program enables visual inertial navigation.
The initial apps for the Helmet include: 1. Augmented step-by-step work instructions; 2. Thermal vision, which can detect temperature issues in equipment or environments; 3. Data visualization – which gives a worker data displays of a type available in a control room; and 4. Connection to remote experts. The company provides working case studies of each of these capabilities operating in actual industrial customer worksites.
Since in the typical case the Helmet must be able to interface with the customer’s existing information ecosystem, this may often require some level of customization of the software or integration. Kammerait states that the company does not plan to offer customization as a service, but will work closely with system integrators, such as Accenture or IBM, who can provide the service.
Regarding connectivity, the Helmet has WiFi, Bluetooth and cellular capability built in. However, he notes that the cellular use has not been certified yet. The exact connectivity methods will typically be dictated by the regulations of the industry that the customer is in.
There are situations where the Helmet may be operated in a mode disconnected from the customer’s ecosystem. An example might be where a party is brought in to a high security facility. In that case, data is collected on the device and transmitted later.
When it is connected, the Helmet can converse with the existing ecosystem. It may be gathering and transmitting new data or receiving existing data. The Helmet will be equipped with information about accepted levels of readings and can recognize anomalous data and send alerts for possible emergencies.
While the Helmet is designed to allow hands-free operation by the wearer to the maximum extent possible, Kammerait stated that in some situations it is necessary to use a secondary device. Currently they can accommodate use of smartphones, but not tablets. In those cases, the secondary device is used for touch control signaling.
The system also enables spoken commands and eye-gaze tracking. If you look to the left for a certain number of seconds, for example, it will bring up information about the item you are looking at.
There has been considerable emphasis on reducing latency of information transfer throughout all of the Helmet’s capabilities. Kammerait says they limit it to a frame and try to achieve a 10 millisecond threshold.
DAQRI’s Journey to The Smart Helmet
DAQRI has had an interesting journey to its present status. Originally the company was a provider of AR software that it sought to sell to companies such as Microsoft and Facebook for wearables apps.
Kammerait explains they found that hardware such as Google Glass and other early stage AR devices were not suited to serious industrial uses. DAQRI started over, “with a blank sheet, where the company wasn’t looking at how to achieve add-on capabilities to consumer devices,” he states.
They determined to focus on industrial enterprises. “We believe we are unique, because even the other providers to the industrial market have some involvement with consumer solutions; we don’t.” He allows that the consumer area “could make sense again” for the company at some point in the future.
During the period that DAQRI was working towards the Helmet, it enjoyed impressive growth, from 30 employees in 2013-14 to over 300 today. The company’s business in the interim was derived in part from pilot programs with leading industrial firms on the issues involved in human-machine interfacing, as well another product line, DAQRI’s 4D Studio. 4D Studio provides an authoring environment for IoT related software, for example, work instructions. Revenues also included a significant services share, such as training users on how to use AR.
Holography & Automotive AR – Two Trees Acquisition
In March 2016 DAQRI acquired Two Trees Photonics, a provider of dynamic holographic technology. After a long development period, Two Trees began supplying technology to the automotive industry for a head up display of information on drivers’ windshields, with Jaguar Land Rover its first customer.
In addition to the opportunities for incorporating holography into DAQRI’s industrial products, Kammerait expresses considerable enthusiasm for the opportunity in the automotive field. “We were somewhat skeptical about the market at first, but this changed after discussions with automotive partner companies.” He sees a large role for AR in vehicles, for apps such as providing cues to drivers, as well as affording 360 information access.
DAQRI Growth Plans & Strategy
Since the company is just at the beginning of the sales cycle for the Helmet, the initial strategy is to drive sales of this product. Kammerait states that they focus on customers who have substantial spend in the areas of plant and equipment monitoring and inspection. In those cases they are probably replacing purchases of tablets or other computing devices. He believes they can achieve sales in the hundreds of thousands of units of the initial Helmet product configuration.
He states that this is a high end capability and they may offer variants at lower price points as well. In any case, he notes that they are already experiencing strong demand and are supply constrained on the Helmet. (CNET speculated that the Helmet will sell for $5,000-15,000 per unit. “Heads-on with a smart helmet that’s named after a delicious drink,” 1/6/16).
He is also optimistic about their potential in the automotive area. He says, “It will be big, like the Helmet.”
In addition to the recent Two Trees acquisition, the company made other acquisitions during 2015, which added to their product line. These included Melon, which offers an EEG-tracking headband, that detects electrical activity of the brain, which, the company stated would have application within the Smart Helmet, as well as potential healthcare uses.
They also acquired ARToolworks, which produced the ARToolKit libraries which DAQRI describes as: “ARToolKit™ is a cross-platform Software Developer Kit (SDK) that allows developers to create AR applications for all major platforms – Linux, Mac, PC, Android and iOS.”
Robotics & AR – The Future
We asked Kammerait about the fact that robotic devices operate in many of the environments that the Helmet aims at, and his view of what the future holds. He stated that he sees both human and machine capabilities improving side-by-side in the industrial setting.
“We’ll see both being optimized,” he said, “but there are situations involving problem solving and critical thinking with many inputs, where human involvement is needed. For more repetitious tasks machines can be better. But a human may be need to guide the machine at critical points.”
He also noted that DAQRI is working on the “bleeding edge of sensor technology.” He says they are looking into developments from a number of areas, including AI, organics, neural networks. He notes that there is a confluence of developments in both the software and physical areas. He points out that “With IoT you need a translation layer for humans to be able to use the information flow.”
Our Take: AR Moves from The Nice to The Important
We started working with and following AR companies about six years ago. We remember the days when the example of an AR app was something like the following: “You’re standing in a piazza in Bologna and you point your smartphone at a restaurant, and there appears on the screen information about the menu, hours of operation, etc.”
Nice to have, in other words, but hardly promising significant valuation creation in the real world.
By contrast, the Smart Helmet is one development that augurs potential real value creation.
Besides noting this highly interesting development, we are very intrigued by the ongoing interplay of machines, i.e., robots, and humans in the workplace. We find Kammerait’s comments about the useful complementary relationships to be worthy of further exploration, which we plan to do, in-depth in the future.
Visit their website: www.daqri.com