AI and Machine Learning

Augmented reality can transform manual labor

As AR-driven spatial computing makes its way into industrial workplace settings through real-time step-by-step instructions and remote mentoring, even the most complex and intricate tasks will no longer require extensive experience to manage and complete them effectively.

By John Tomizuka May 25, 2021
How navigation used to work (left) compared to how it works now (right) thanks to current technology. Courtesy: Taqtile

The COVID-19 pandemic has accelerated the adoption of digital transformation strategies in many industries, ushering in foundational changes that likely would have taken five to 10 years to occur otherwise, but there are many areas that have remained on the sidelines of this revolution. Consider the dozens of industries reliant on skilled manual labor.

While organizations may have data on when a task was started and when it was completed through existing work management software, traditionally, there has been no way to measure and track overall efficiency and effectiveness in industrial workforce settings. However, we’re on the cusp of tremendous change, thanks to augmented reality (AR), artificial intelligence (AI) and the dawn of spatial computing.

To better understand what this new world will look like, consider the evolution of maps. Let’s say you need to find a restaurant. Not long ago, this task that might take all of five seconds today, was an arduous process that required a combination of planning with physical maps or written directions based on a mix of street names and landmarks, finding coordinates in indexes and stopping to make calls if you got lost (see Figure 1). Then you had to reverse-engineer that journey to get home. Today’s process could not be more different:

Today, all a person has to do is say, “Find a good burger nearby,” and the AI within the smartphone will produce many possible answers. So, 10 time-consuming steps have been condensed into one simple statement (see Figure 2).

Figure 1: The Thomas Guide coordinate system. Courtesy: Taqtile

Figure 1: The Thomas Guide coordinate system. Courtesy: Taqtile

Industrial context

Can we apply this divide in the way people navigate from place to place to the industrial workforce and the processes they use to get their jobs done? Can we create a hyper-localized “navigation” of what to do and how to do it for the industrial workforce? In manual labor settings like factories and other industrial environments, if there is an emergency with a system or piece of machinery and the right expert is not onsite to fix it, troubleshooting the problem often means a reliance on physical manuals to address the issue. It’s the Thomas Guide in the form of standard operating procedures. These tasks are much harder than driving a car and require the user to be aware of what is happening across many different subsystems, as well as an active understanding of how they interact with each other.

As AR-driven spatial computing via heads-up display increasingly makes its way into these settings through real-time step-by-step instructions and remote mentoring, even the most complex and intricate tasks will no longer require extensive experience to manage and complete them. Much of the time spent troubleshooting and addressing problems in complex industrial systems today is wasted trying to figure out what is wrong in the first place. The expansive data generated through this transformation will open the door for AI to revolutionize the diagnostics process.

Figure 2: Modern navigation system. Courtesy: Taqtile

Figure 2: Modern navigation system. Courtesy: Taqtile

With widespread use of spatial computing, organizations will have data on everything happening within these environments. Combined with industrial internet of things (IIoT) data, 5G networks with powerful edge-computing capabilities, and the incorporation of sensors into legacy equipment, organizations will be poised for a massive leap forward in understanding a wide range of complex, manual work. In time, we will have systems capable of diagnosing problems and leading even novice users through the process of troubleshooting complex challenges.

Industrial environments are still working off a Thomas Guide model, but just as manual driving and written navigation have given way to cars capable of assisting drivers with important safety information and navigating them based on real-time traffic data, skilled manual labor, too, will evolve in ways that will add simplicity, drive more intuitive processes, and streamline productivity functionally across organizations and industries.


John Tomizuka
Author Bio: John Tomizuka is CTO and co-founder of Taqtile, maker of the Manifest AR work-instruction platform.