Nov 27, 2018
Augmented Reality (AR) is the latest bandwagon to grip the minds of the public. There has been a plethora of AR announcements over the last year - including Mozilla’s Firefox Reality (an AR/VR browser) and Snapchat’s Lense Studio that lets users create their own filters. The tech giants have all released AR developments tools (Apple’s ARkit, Facebook’s Spark AR Studio and Google’s ARCore) and announced plans to launch AR glasses.
Such heavy investment in the technology indicates real potential beyond entertainment, and there are some people looking deeper into AR and how to combine it with other emerging technologies. I spoke to three individuals, Daniel Spruce, Ryan Hooks and Adrian Leu, who are proving that AR is useful for far more than just Pokémon Go.
Farming 4.0
Augmented reality is a great way to visualize complex ideas, like what all that code actually does, or how to cultivate plants in restricted conditions. One company with a view of a technological future is Plant Vision, who are aiming to create a decentralized ‘collaborative’ AI ledger for plant breeding & optimization.
Using AR, each ‘master grower’ receives equity in the footage they annotate. Recorded in the RGB color spectrum, infrared for ‘early disease detection’ and ultraviolet for ‘flowering and pollination’, the growers then add this data to a digital ledger to collaboratively train the AI system, and they receive value over time as their data is put into commercial use.
Considering the company’s roots in the rapidly growing cannabis industry and the complex, valuable data that comes with it - ‘the dataset just for powdery mildew for cannabis will be worth billions a year’ - the ambitions of Plant Vision are certainly not stuck in the mud.
Transferring experience from his first companies Isabel/Huxley, founder Ryan Hooks hopes to harness AR and AI to increase yields and motivate farmers around the world. “For vine crops like cannabis, tomatoes and cucumbers, a Dutch grower can get 500% more per square meter just by knowing how to take care of the plant, so augmentation has so much potential,” Ryan says.
Based at Wageningen University, the project is currently focused on making tools for plant scientists in the Netherlands. With expansion planned in 2019, Hooks intends to ‘bring multi-million dollar Digital Phenotyping into our pockets’ thanks to high-powered GPUs and Machine Learning capabilities inside the latest smartphones. (Read: Microsoft, Tencent, and Intel grow autonomously at Wageningen)
“It’s not AR for AR’s sake, it’s a hands-free system for understanding plants via AI,” says Hooks, “I see augmentation as a bridge to automation, as it will take about 10-20 years for robots to be affordable in many plant sectors - instead of a million dollar robot you could have 50+ augmented growers.”
Beyond the bandwagon
While the visual media industry may be getting a reputation for blatantly capitalising on AR, there is a large amount of research behind closed doors aiming to advance AR technology in general. Adrian Leu, CEO of Inition, talked to me about where AR is headed and the investment that is happening in the visual media industry to solve limitations of this technology. “Current headsets, like Magic Leap or Hololens, are concerned with one major aspect which is to convincingly align and display virtual objects onto the physical world. However, there is no cognitive correlation at the moment between those objects and the physical world.”
Leu calls this ‘ability to process the surrounding world’ contextual intelligence and argues that this represents the next stage of AR that can gather information from its environment and learn to ‘see’ and display more appropriate information.
“One can imagine soon how a certain wearable could scan an environment and process it in the cloud,” says Leu, combining AR with advanced AI to increase this contextual relevance by ‘feeding data from sensors that study how the physical world behaves and evolves.’ This kind of intelligent AR would be a more hardware-based route towards achieving real computer vision, a cornerstone of AI research, and ‘could have a major impact in applications like training, remote maintenance, and collaborative design.’
Augmented Reality may be the latest craze for the visual entertainment industry, and an exciting new technology for the tech giants to sink their teeth into, but as is often the case there are far more interesting things going on in the background. Whether used in individual projects to make programming more accessible, ambitious plant-rearing collaborative AI, or in promising research behind the silver screen, AR has a bold future ahead of it.