Generative AI: commodity or core business?
04 March 2024
The choice between buying or developing generative AI in-house depends on the use, the data and the competitive advantage sought. Eleven proposes a methodology for…
Innovation
In 2013 Google released its first “Google Glasses”, a pioneering product that brought augmented reality experiences to the masses. However, Google Glasses were quickly dismissed as a gimmick without much use and did not meet the expected success. Production almost stopped after two years.
Almost 10 years later, the development of Google Glass continues, but only in a limited version for businesses. As it happens, this is where the most immediate applications of augmented reality lie, and more particularly in the industry through the optimization of industrial processes. As machines become more and more complex, like cars that integrate more and more electronics and sensors, the demand for qualified technicians is hard to meet. Assistance to increase the productivity of existing employees or to quickly train new recruits can therefore respond to a very pressing need. It is in this sense that augmented reality can be used in the industry.
Augmented reality (AR) superimposes information, texts, images or sounds, on the world as we perceive it. It is then possible to interact with the environment using 2D or 3D animations and computer-generated images for entertainment or assistance.
Unlike virtual reality experiences, which are totally immersive in an artificial world, the adoption of augmented reality tools seems to be easier, socially and physically. Indeed, augmented reality does not cause the headache or nausea problems encountered during the first VR experiences, especially in the video game industry.
It was the game Pokemon Go that democratized the use of AR in 2016 with over 700 million downloads. This mobile application uses GPS information and the user’s phone camera as an indicator to bring up Pokemons to capture in streets, parks and monuments around the world.
Other industries such as luxury and cosmetics have decided to invest in augmented reality to transform the consumer experience. From the LVMH group to L’Oréal, they have since bet on diversifying their customer experience through digital applications using AR. For example, Modiface allows customers to virtually test Garnier products from their own cell phones.
Although the use of augmented reality is still mainly focused on transforming the user experience, many R&D projects tend to diversify these uses, particularly in the industrial context of maintenance and production line optimization.
The applications of augmented reality to industry mainly concern three areas:
Overall, it looks like the “Industry 4.0” will be an augmented industry and a large number of use cases remain to be explored. However, there are still some very real limitations to completely industrializing certain applications.
The desire to merge the real and virtual worlds in environments as complex as (e.g.) a construction site will only be possible after many challenges have been overcome.
First of all, artificial intelligence (AI) must perfectly recognize the elements around the user: Obstacles, buildings or movements and facial expressions. This issue comes before the creation of graphic content and is one of the major problems of artificial intelligence today. Image processing – or computer vision – is a branch of data science in full expansion, which aims to analyze images in order to extract information that is currently reserved for human understanding. Many computer vision applications are now industrialized but the compatibility of such solutions with augmented reality environments remains limited. Many initiatives such as CoreML or TF mobile launched by Google are encouraging and suggest that in a few years, AI and AR will develop jointly. In the meantime, this remains a limiting factor for diversifying the use cases that can be easily industrialized today.
A second challenge is the availability of a digitized environment. Drawings, “digital twins”, processes, operating procedures, etc. must be available with the right level of detail and quality to be able to be overlaid on reality. Again, AI can help to digitize, interpret, make available and verify existing content, however a significant effort must be made to initiate and maintain such documentation.
In addition, the performance of the embedded hardware on cell phones can be limiting. A GPS, an accelerometer and a gyroscope are essential to accurately place objects in space. The introduction of Lidar sensors in iPhone 12 models has considerably increased the measurement of relative distances between objects but is still reserved for a very limited range of mobile devices.
Finally, the use of connected and augmented reality tools requires a change management phase with operational staff. The user experience and business issues remain the most important elements when defining use cases. In the case of complex tools – connected glasses or mobile applications in an industrial context – the team of technicians must be supported to facilitate their adoption.
To face these challenges, eleven strategy has helped a world leader in industrial installations to create a mobile application allowing experts and non-experts to perform maintenance operations on fiber optic networks in a faster way and with more reliable data directly accessible on the field. Our client was thus able to create a competitive offer through a native solution and better position itself on tenders in a dense market.
Ariel Cohen Codar, Simon Georges-Kot, Morand Studer
Share
Sur le même sujet