History of Datch
I began my career as an engineer working with Australian Navy submarines, before moving to the UK to work in aircraft systems design and manufacturing. I founded Datch in 2018 with my cofounders Aric Thorn and Ben Purcell, who both had backgrounds in the energy sector in New Zealand, UK, and Germany. Throughout our career journeys, we all came across the same limitations: while the aircraft we were building and the energy projects we were managing were highly advanced, the tools and processes for managing the everyday work was stuck at least 20 years in the past.
Manufacturers, airlines, and airports all have one thing in common: they require knowledge data coming in from their workforces, consistently and accurately, while also balancing tool-time requirements, compliance, and data quality. Updating work orders, recording inspections, and completing safety reports and audits require time and effort, but with 400 million people now regularly engaging with voice tools in their day-to-day routines, voice AI has emerged as an elegant solution for solving this“input problem” for the frontline workforce.
Datch has come a long way in the last few years, working with customers across the manufacturing, mining, and energy sectors, and we recently announced our entry into aerospace with a strategic investment from Boeing.
Digital Transformation of Aerospace
There’s a lot of activity in this space and a long way to go before the industry achieves ubiquity, but there are a number of emerging solutions such as Datch, Realwear, and Mira that are putting together the pieces of that puzzle.
Organizations that are well run have a curse - the more diligent they are at collecting information and processing that information, the less time they get to spend on hands-on, ‘value add’ work. There was a study that came out a few years ago by McKinsey that found that across industry, 22% of all hours worked was considered “recoverable time” spent on these business processes. This is really where voice-visual technology comes in. For example, at Datch we use verbal conversations, which are the most natural form of human communication, to help with these repetitive, compulsory but critical, business processes. Other companies (like Realwear) are providing the hardware for hands–free work management, and Mira is solving this through clever heads-up displays and check-listing workflows.
And what is meant by processes? This is wherever humans have to retain or retrieve knowledge at the frontline. If you're an engineer or technician, these might be work orders, tasks cards, tech logs, and maintenance logs. In quality and safety, these might be things like incident reporting, SOPs, 5-why procedures, or audits. Ultimately there are hundreds of these processes being carried out every day by every person throughout an organization. There are a few companies tackling these problems today with a variety of solutions, but it requires a radical shift in thinking and reconfiguration of both software and hardware layers.
This illustration is showing the classical approach to knowledge acquisition at the frontline, and also for information recall. Typically, workers and engineers need to use a combination of either paper forms or desktop terminals (often both) in order to record information against their jobs, to receive instructions on their jobs, or to troubleshoot their work. This is done through “asynchronous” communication and requires the worker to stop what they’re doing in order to record or query data. This results in either a loss of data quality or a loss of time. Most often, it tends to be both, due to the trade-offs that need to be made.
The shift to voice-visual is a shift towards real-time, synchronous communication. In the future, whenever there is a process event, workers simply need to declare it. If workers need to troubleshoot, they simply need to ask a question. If workers need to record knowledge data throughout a job, they simply need to talk through it. If workers need access to information (such as an assembly drawing or a set of work instructions), they have it on demand. The digital transformation to voice-visual is mostly concentrated on the creation of a synchronous, frictionless flow of information.
Human language is incredibly good at compressing data without losing quality. Natural Language Processing (NLP) mimics the human brain, not only in understanding the words that are said, but also in understanding the intent and relaying that conversation as structured data. Artificial Intelligence (AI) becomes that translation layer between the efficient, human-side and the structured, systematic, machine/company-level needs.
The voice-visual approach requires both a hardware and a software layer to automate the synchronous flow of information between the frontline workers and the systems of record.
- On the hardware side, audio must be captured cleanly, provide audio output, and images/videos must also be captured while providing visual feedback on many processes.
- On the software side, the raw input data streams must run through machine learning algorithms in order to extract intent, context, and general structure to that data, before routing it to an ERP system. The technology must understand what information to send back to the user, and where it needs to pull it from.
The future of frontline work is where you no longer need to worry about where you are or what you need to do in order to do your job well, in a way that satisfies your needs, as well as your company’s needs. This has historically been a difficult trade-off, but the ultimate purpose of voice-visual is to eliminate any sacrifices.
The voice-visual industry is young, but growing very quickly. At Datch, we’re always looking to learn about new applications and new ways to leverage these tools to improve workflows.
We were recently featured at "Aerospace & Defense 4.0 - What It Is & Why It Matters" presented by AeroTech Review. Watch the playback here.
Interested in learning more about Datch? Request a demo.