July 03, 2019 by Tyler Charboneau
AI has also emerged as a key player in the engineering realm. Businesses are increasingly pursuing automation, which is AI’s calling card. Such algorithm-based cognitive tasks can be time-consuming for humans. AI development constitutes an exciting new engineering frontier to explore as technology evolves. Both AR and AI are transforming the global productivity landscape.
In fact, we’ve previously touched on the potential that AR and VR have to revolutionise workplace productivity. But while VR holds immense promise, it's AR that has stormed ahead at a breakneck pace thanks to existing hardware permeation. Accordingly, a large proportion of devices currently satisfy AR’s technical requirements.
Both desktop and mobile are primed for AR development. Most enterprise players have engineering teams and resources dedicated to hardware and software. These same companies maintain vast networks of computers and smartphones. Development and engineering become simpler when hardware is largely ubiquitous. Many AR capabilities on desktop are hardware enabled, yet handcuffed by a stationary platform. In these instances, AR capabilities are confined to productivity apps and browsers.
Mobile devices hold incredible promise in the AR space. While computers are traditional powerhouse platforms, mobile devices ship with hardware the components paramount to immersive AR experiences: sensors. Gyroscopes, accelerometers, depth and proximity sensors, and more fuel real-time augmented experiences. Each component has power and processing demands. Manufacturers must, therefore, ensure robust experiences are battery efficient and lightweight.
AR allows us to explore new digital possibilities. Image courtesy of Unsplash.
We have observed these goals manifesting themselves in the Google Glass Enterprise Edition 2, which is carving a growing industrial niche. Power demands are crucial: we don’t want to design AR solutions that are only usable for a few hours when a typical shift can last over nine. Since compact electronics introduce limitations, these programs must operate effectively with fewer resources.
Consequently, designing an AR experience can be challenging, even for well-versed electrical engineers. Software and hardware teams must collaborate more than ever to ensure features are functionally sound. This is true regardless of industry.
Speaking of industries, professionals hold AR in fairly high regard.
In a joint March 2019 survey by Perkins Coie and the XR Association, 70 per cent of enterprise respondents believed the AR market will surpass the VR market in revenue. Additionally, 81 per cent of those respondents “predicted it would happen within five years.” These figures demonstrate, at the very least, heightened confidence in AR’s future.
When asked which industry will see the most AR/VR/MR (mixed reality) investment, survey respondents answered proportionally:
A percentage-based graph that shows the results of a survey that asks respondents which industries they think will see the most investment over the next 12 months.
Note that these percentages include AR, VR, and MR together, which may skew the data somewhat. However, we know that gaming, healthcare, and marketing have immense potential in the AR. We have previously discussed advancements in AR medical training and potential surgical applications pertaining to anatomy. Brands like Amazon, Angry Orchard, and Apple incorporate AR into their marketing strategies, leveraging your smartphone camera to perform live rendering and even offer garment fitting. As we’ve discussed briefly before, Web AR allows consumers to interact with products while browsing.
These features rely upon a given device’s hardware, leaving engineers and developers with the task of ensuring compatibility. It’s also worth mentioning that AR’s won’t exist in a vacuum. The technology can pair with VR to form richer applications, like in MR’s case.
The manufacturing category also encompasses the automotive world. We’ve previously seen rudimentary forms of AR in the early to mid-2000s with heads-up displays (HUDs). Since then, these HUDs have included maps, directions, and other information to keep drivers focused on the road ahead. Advanced awareness features rely on increasingly-complex sensors to work properly. These intricate electronics rely on specialised component designs based on technological needs and space constraints.
As hardware improves, it will support a greater range of functionality. This software bridge is key in making complex renderings and AR features work effectively without delay. Silicon is shrinking almost annually while we move from a 14-nanometre process down to 10 nanometres, now down to 7, and so forth. Increasing transistor density and form-factor reductions have made AR advancements possible. Further changes will help AR-enabled devices become less cumbersome.
The hottest new technology on the block, AI seeks to drive process improvement for businesses across many industries. Companies are joining the fray at a rapid pace. However, many businesses still have reservations with AI based on common misconceptions. While AI seems complicated, it’s actually quite a bit easier to implement than teams may expect. Not only that, algorithms are becoming easier to engineer as AI matures.
According to a report by PricewaterhouseCoopers (PwC) from earlier this year, scalability will determine AI’s success in the next decade. The piece also claims the existing quantity of AI algorithms is relatively small. While we envision AI as the wild west of technological innovation, the market is tamer than previously believed. Fewer options make it easier to become literate in these technologies. As teams grow more knowledgeable, algorithm engineering will be less daunting and training will be feasible.
According to PwC, incorporating AI into your business model begins with one core algorithm. Companies can automate one element of their business as a proof of concept. Once these changes make positive impacts, expansion is possible. This is where AI shines. Algorithms are shockingly adaptable and easily configured to tackle changing business needs. These adaptations are exciting new avenues for AI engineers, who must determine how to squeeze maximum potential from every program.
As expected, companies surveyed by PwC have different plans for adopting AI:
The results of PricewaterhouseCoopers' (PwC) artificial intelligence implementation survey. Image courtesy of PwC.
To prepare for AI implementation, businesses in multiple industries must assemble teams built around AI specialists. This will help AI initiatives get off the ground more effectively.
The report outlines some key industries currently capitalising on AI’s productivity promise.
Though businesses will leverage AI, most economic impacts will be felt on the consumption side. Since AI evolves based on usage patterns, companies will use it to provide personalised services. These will likely focus on contextually-relevant data and products.
In PwC’s estimation, there are over 300 use cases for AI today. Many of these are rooted in the healthcare, retail, and automotive realms. Imagine the possibilities regarding patient care, product recommendations, and more. The financial sector and payroll divisions worldwide will benefit from automated invoicing, taxation, and processing. Engineers will help lead this productivity charge while employing hardware-conscious approaches.
AI is also adept at highlighting trends across many verticals. Artificial intelligence will facilitate sound decision making pertaining to business strategy.
That said, AI will become more accessible at home, in the office, and in our pockets. Akin to AR, AI is driven by capable hardware. Since neural networks behind the technology can be incredibly intricate, processors must be up to the task.
Luckily, much of our current mobile technology is future-proofed to support complex AI computations. Apple’s A12 Bionic chip, for example, has a Neural Engine powering up to 5 trillion operations per second. The device can easily handle any AI tasks thrown its way. Similarly, Huawei’s Kirin 970 processor has over 5.5 million transistors. It’s designed from the ground up to tackle AI operations, solidifying Huawei as a mobile AI leader.
From left to right: Apple A12 Bionic chip and Huawei’s Kirin 970 processor. Image courtesy of Wikipedia and Gigazine.
As the mobile space innovates, semiconductor manufacturers like TSMC and Qualcomm must evolve in lockstep. Just like the medical field will always need doctors, the hardware space will require a growing number of electrical engineers to bridge the gap between possibility and reality.
As AI becomes more intelligent, further questions will also arise regarding ethics and safeguards. AI development must occur in a responsible manner. These concerns relate to transparency, security, biases, legalities, and governance.
Both AR and AI will quickly change how we do business. They are thrilling new technological frontiers that have largely been unexplored. Their maturity, though five to ten years away, will unlock powerful new productivity shortcuts for our most-demanding global industries.
Most exciting of all, the hardware is already there. We need an enthusiastic stable of developers and engineers to help the technology reach saturation.
The tech industry can improve countless lives via user experiences and automation—collectively, we’re sitting with anticipation on the edges of our seats.
