My engineering background and CAD experience led to my work with intelligent assets at Infosys. Infosys is a multinational IT services company, focused on augmenting human potential across industries with the assistance of AI. As a Senior Experience Design Lead at the office of the CEO, I directly supported the company's initiative by creating rapid prototypes to engage customers and spur discussions around industry testable use-cases for emerging technology.
Intelligent assets are large industrial assets that are densely sensored, can be monitored and interactected with remotely, and viewed locally with more information than physically available. During a design thinking session with our engineering services team around how to convey this concept and Infosys' offerings to clients, I drew a connection between an open source 3D printable model I had previously stumbled upon online, microcontrollers I was using for a Seeing Spaces project, and the Microsoft Hololens developer kit our team had recently received. I crafted a storyboard of how these components could be combined to create an experience that showcased the company's capabilities across different departments. I initially encountered some resistance with the small upfront costs associated with the project, a common and frustrating issue I've found working in large corporate companies. I truly believed in the idea and proceeded by covering the ~$300 of expenses myself.
I was working on an extremely short timeline, less than fifteen business days to deliver a fully functional experience from a storyboard. I reached out to our AR/VR team for their expertise on the most efficient path forward and enlisted two of their talented team members to help - one 3D artist and one Unity Developer. I made a few modifications to the 3D model to fit it with off-the-shelf DC motors and began the multi-day print.
While this was happening, I met with experts in aero engineering and artificial intelligence to get a feel for what components must be present in this experience to remain technologically compelling. With these findings in mind, I identified sensors that could loosely model what similar sensors would capture on a full scale turbofan. I setup a rudimentary gateway to collect sensor data and send it to a cloud database. I learned how to create an API that would make this data stream accessible by our Hololens application. Seeing the first bits of data flow into the headset was so satisfying, I forgot about cursing web sockets moments prior.
Our team worked through the night, prior to me flying to Columbus, Ohio with our final creation. I presented the experience with our engineering services team at GE's Aviation Conference and to the CEO of GE Digital. It was so rewarding to hear the positive reviews that it made my sleep deficit bearable. As cliche as it sounds, this experience was a strong reminder to take risks and follow through on gut-insticts. Following this event, I was asked to present the demo around the world at corporate events, at the World Economic Forum in Davos, Switzerland, and to other industry leaders.
I scaled the demo by creating a set of instructions, a bill of materials, and marketing material for distribution across our organization. The internal publicity afforded me the opportunity to help setup innovation hubs in Germany, China, India and Australia, to foster the creation of similar experiences. This demo became the most widely used demo by the 200K+ person organization of Infosys in 2016, and it is still being shown today. Building something for this magnitude of people to use and experience was a gratifying experience and a realization of how visions can propagate.
The Pitch:
This 2-week sprint design was created to illustrate an Artificial Intelligence IoT use case. A 3D-printed open-sourced aircraft turbine model is used to represent a real, large, complex asset. The model was fitted with motors and sensors that stream data to the cloud where aggregation and analysis is performed. Augmented reality enables users to visualize real-time data overlaid on the physical turbine and forecasted outcomes on the turbine’s digital twin. Users can have a spoken conversation with the engine to understand its current status and health. The AI that enable this chat function references an ontology map of the engine, produced from the service manuals of real gas turbine engines. In practice, such insights allow manufactures and service providers of large industrial assets to transition from routine maintenance to high fidelity condition based maintenance and reduce down time of the asset. An enhanced user experience with AR also offers opportunities upstream to contextually visualize bench test results during development, and downstream to augment the capabilities of field technicians during repair.
