The AR Industry’s Continuing Growing Pains Are Necessary

Trying to push past the limits of current engineering obstacles in augmented reality

Augmented reality applications are limited by the tools that developers use to build them. In any field, people are limited by the capabilities of the tools they possess. The tools you use make a difference in the quality and scope of your work, and limit what you can build and accomplish.

Augmented reality applications for industrial uses such as those in manufacturing—whether it’s to bolster complex assembly, support maintenance or repair technicians, or for quality assurance—are generally reported on as being in the experimental phase.  

There haven’t really been any true breakthrough industrial applications in augmented reality, partially because the technology is new, and partly because the tools that software developers are limited in scope.

Perhaps combining elements of data science and artificial intelligence with augmented reality is a good idea, but developers may feel like it’s difficult to find an entryway or a foothold to begin incorporating AI into augmented reality applications. An augmented reality application could perhaps be greatly served by incorporating AI into it. AI would allow AR to interact with physical environments on a multidimensional level by incorporating real world object tagging, for example. But the augmented reality market seems to be going through some serious growing pains this year. One promising collaboration involving AI and augmented reality has already been extinguished.

IBM and Unity’s AI SDK for Augmented Reality Applications Has Been Quietly Discontinued

Earlier this year, Unity announced a partnership with IBM to launch a new IBM Watson Unity SDK, a programming interface that gives developers a way to add cloud-based AI to their applications. (Image courtesy of IBM.)

Earlier this year, Unity announced a partnership with IBM to launch a new IBM Watson Unity SDK, a programming interface that gives developers a way to add cloud-based AI to their applications. (Image courtesy of IBM.)

From IBM, the perspective is data-centric. Huge swaths of data are being generated at an enormous rate now—2.5 quintillion bytes of data per day—and IBM Research writes that 90 percent of the data that exists today was created over only the last two years. This is way more data than can be properly analyzed or utilized in a productive way.

So IBM was looking to improve data visualization methods using augmented reality driven visual environments. An IBM project called Immersive Insights was tackling this issue by giving AR developers a chance to incorporate computer vision and speech recognition AI elements into Unity applications. IBM was also offering developers the ability to incorporate speech recognition with classification and language translation features in the IBM Watson Unity SDK. Not a bad idea.

Computer vision is the part of AI that is seeping into augmented reality applications built by companies like Vuforia, Unity, Microsoft and Blippar. IBM has Vision API for computer vision that was available on the IBM Watson Unity SDK.

Unfortunately, the Unity and IBM Watson Unity SDK has been discontinued, and the reasons are unknown at this point. But it points to a larger trend of cutting the fat in AR.

The AR Industry is Coming Back to Reality

Exhausted hype, overpromising, underdelivering and a reaction to market volatility between China and the US hit the augmented reality market and a wave of prominent startups went kaput earlier this year. Meta Company, who sells the Meta 2 AR headset became insolvent.

Osterhaut Design Group (ODG), whose founder Ralph Osterhaut sold AR-related patents to Microsoft pre-HoloLens is also bankrupt. UK-based Blippar, who were on a mission to create a Google search engine for image-based queries entered a form of bankruptcy. Magic Leap, HoloLens and DAQRI now stand out as most healthy. Magic Leap has tons of funding, Microsoft has billions and a recent defense contract win. Industrial augmented reality company DAQRI has been extremely quiet for about a year now.

Maybe a problem is the lack of real industrial applications?

Industrial applications like Honeywell’s cloud-based simulation tool that leverages a unique combination of augmented reality (AR) and virtual reality (VR) to train industrial plant personnel works partially because it fills an urgent and practical need: Nearly 50 percent of industrial plant personnel are going to retire in five years. The cloud-based AR and VR simulation tool is a product of something called Honeywell Connected Plant Skills Insight Immersive Competency, which was created to speed up training and the transfer of knowledge to new industrial workers. Honeywell’s HoloLens solution leverages the cloud. (Image courtesy of Honeywell.)

Honeywell's new HoloLens and Windows Mixed Reality training solution leverages the cloud to provide simulations of their C300 controller to help train new industrial employees. (Image courtesy of Honeywell.)

Honeywell’s new HoloLens and Windows Mixed Reality training solution leverages the cloud to provide simulations of their C300 controller to help train new industrial employees. (Image courtesy of Honeywell.)

Honeywell has collected a massive amount of data from different scenarios involving the C300 over the years, and this cloud-based simulation tool allows trainees to wear a Microsoft HoloLens or Windows Mixed Reality headset to experience various workplace situations, including cable failures, power supply failure and switchovers. The immersive training is used to test, teach and sharpen the budding skills of new personnel.

The trainee will have to identify the faulty C300 controller, dismount it from the IOTA adapter, mount a new C300 and power it on for the alarm to stop in the Experion Station. The virtual environment for each specific job activity simulation accessed by trainers and trainees is hosted on a private and secure (as the cloud can be) cloud server. This novel tool also allows plant managers to compare the competency of their staff to the performance of their plant. According to Honeywell, this simulation tool shows up to a 100 percent improvement in outcomes versus traditional training methods and reduces the time it takes to train a new employee by up to 66 percent. Industrial applications like these are still working and worth paying attention to.

If only professional grade augmented reality applications for enterprise and industrial users proved useful enough to offset the cost of implementing the technology at a large scale. Another promising example is the DAQRI Worksense suite of applications, which could still be poised to demonstrate value in cost-benefit analyses from a range of industrial and enterprise organizations without cumbersome issues that prevent implementation. If there’s value in the adoption of AR tech like the DAQRI Smart Glasses, then their success will become a no-brainer. 

And the ability to 3D scan work environments and view BIM models in-situ is promising. For example, since DAQRI’s current customers are all still privy to receive a trial of DAQRI Worksense Pro (no word on the promised Worksense Standard yet), decision-makers will be able to discern what is working and what is not, making adjustments along the way. (Image courtesy of DAQRI.)

And the ability to 3D scan work environments and view BIM models in-situ is promising. For example, since DAQRI’s current customers are all still privy to receive a trial of DAQRI Worksense Pro (no word on the promised Worksense Standard yet), decision-makers will be able to discern what is working and what is not, making adjustments along the way. (Image courtesy of DAQRI.)

Customers like Norwegian-based technology company SHM use DAQRI Smart Helmets and Smart Glasses for maritime, offshore and aquaculture industry activities. They provide solutions and services that cover broad areas like cyber-physical systems, maritime technology, computer science and automation as well as aquaculture. They use DAQRI technology as prime factors in cutting costs of services they provide. For example, aquaculture users can video conference with remote experts in-situ and without the use of their hands.

But DAQRI’s been really quiet for a while now. There’s been no new entries on their blog for almost a year. Time will tell.

Bottom Line

AR is going through growing pains, and perhaps the industry was taken off-guard by the American-Chinese trade wars which clearly affected a few key relationships for promising organizations. HoloLens is still the big winner of a huge US government contract, but Magic Leap is still cooking with venture capital funding. As far as industrial engineering applications, DAQRI is the most promising startup but the persistent radio silence is probably becoming alarming for those with a vested interest. HoloLens and Honeywell’s collaboration seems a better fit than Unity and IBM, but we’ll see where things go as the industry matures a little and continues to unfold.