The Rolls Royce Aviation, Indianapolis, US, manufactures nozzles for helicopter turbine engines.
Currently the welding department manually tracks the non-scheduled parts(parts like nuts and bolts whose quantities cannot be defined, but are essential for the manufacturing process).This is prone to errors and delays the timely manufacturing and delivery of the engines and kits.
As part of the solution, we designed an automated digital tracking system using internet of things (IOT) that provides real time information about the inventory status and order updates anytime anywhere.
Watch the hardware and the corresponding UI in action.
write something here
The shelf as indicated below had a slanting partition to hold to bins. Bins were of varying sizes with the parts in it uniquely identified by the part number.
We talked to the stakeholders, Ryan(supervisor of the welding department), and welders to understand their pain points and needs. Further stakeholders clarified the business goals and constraints we would need to work with.
Formative research helped us get a good idea of how inventory management is done at small-scale and large-scale companies, and automation is used in existing similar systems.
Some insights that we gathered were:
Load cells were the most commonly used sensor on bins to calculate the weight. Analytics helped decide the rate of parts consumption based on usage habits and patterns. RFIDs were widely used to track orders and parts once out for delivery. One of the existing technologies that really inspired me was the BinSense bin management system.
Post the research phase, we collectively discussed and exchanged thoughts from our individual findings to put together the existing inventory management process at Rolls Royce.
From the above workflow we were able to gather and categorize different insights as below:
*These insights helped us give a good direction of what we might need in terms of sensors, technology and infrastructure going into the ideation phase.
Ryan(supervisor) and the welders identified the parts by their numbers(parts ID) and not the bins.
Welders did not care to put the bins back to its place and used it at their convenience causing unorganised bin placement.
Ryan had to manually collect the empty bins lying around the station.
Ryan placed the empty bins at the top of the shelf so that it could be pickup up by the dock worker once the restocking order was placed.
Welders were mostly goverened by the union workforce and hence did not care about the bins or the parts.
welders would not take any additional responsiblity of organizing the shelf as it was considered additional work for them.
"The lack of communication when bins get too empty is the most frustrating part. It's even more frustrating when the supplier is also out of those parts which causes extremely long lead times."
The lighting in the station was dull and welders could not see the bins from a distance.
The docking station where the parts arrived was pretty far from the shelf.
There was no power source near to the shelf.
Now was the time to take all the analysis and learnings and empathise with the users.
The team came together post the research phase to collaboratively empathise with the users, empathy maps helped us understand the pain points and goals of the users as a team so that everyone in the team is on the same page throughout the design process. Further we created a persona for the team to empathise with the user throughout the future phases of the design process.(to keep the user in mind always while designing)
We defined our problem statement, constraints and design requirements.
How might we help Rolls-Royce Aviation ensure continuous availability of parts needed to manufacture airplane nozzles in a timely manner using internet of things and eliminate the burden of manual bin management?
After understanding the goals of the business, it was important to also understand limitations that we would need to work with. Some constraints or boundaries that would help us not become too ambitious:
After understanding the goals of the business and user pain points we defined the design requirements that we would want to achieve through our solution so that it satisfies the needs of the personas we created and alleviates the pains of our users:
We ideated to come up with different solution possibilities and gathered feedback based on which we iterated and designed our final solution
For the ideation phase, we identified different permutations and combinations of load sensors, IR sensors, computer vision and other technologies that we could use in our solution:
For the ideation phase, we identified different permutations and combinations that can be used to improve the inventory management process:
Below are some feedback from the clients:
The clients were happy with the provided solution as it ticked most of the boxes that were defined as part of the design requirements.
However, one feedback that we wanted to iterate on was that the part of the workflow where the refilled stock had to be brought back to shelf still involved some manual work and there was scope for improvement
Based on the client feedback, we regrouped and iterated on the solution to update a few parts of it and proposed the final solution.
Based on the client feedback we modified the solution a bit and added one new set of bins considered as temporary bins that would be only used to restock empty bins at the dock , this also helped to reduce the delivery lead times.
Based on the client feedback we modified the solution a bit and added one new set of bins considered as temporary bins that would be only used to restock empty bins at the dock , this also helped to reduce the delivery lead times.
Below is a proof of concept video that we created for better understand of how the UI and the hardware component are mapped and interact with each other.
Easy access to parts inventory anywhere, anytime.
Parts shortage and bin misplaced alerts.
Order history and prioritisation
Easy visual indication
order easily through the application.
track order and prioritise them.
Informed decision making.
understand usage patterns.
Here are a few of more of my case studies.
