Scroll to top

ITrack: Implementing Auto Inventory Tracking For Rolls-Royce

ITrack: Implementing Auto Inventory Tracking For Rolls-Royce

Enterprise UI

User Experience

IoT Design

AI/ML

+30%

Improved task completion rate, achieved by streamlining process touch-points from 4 -> 1 for Prime persona.

-45%

Reduce in time required for users to perform inventory tasks such as checking, creating orders, communicating requirements.

Introduction

Rolls Royce facility located in Indianapolis manufactures aerospace engines and parts. The facility has an automated SAP system to track their inventory which is crucial to their production cycles. However their Welding department has issues with tracking the non schedulable parts, which can’t be managed with through existing SAP system and hence the term non schedulable. Our team set out to innovate an end-to-end IoT solution aided by a mobile-web platform to streamline this process.

MY ROLE

Research, analysis, conceptualization, information architecture, designing high-fidelity prototypes and evaluation.

TEAM

3 Product Designer

1 Developer

1 Project Manager

3 Stakeholders

SOFTWARE

DURATION

8 weeks

Proof of Concept

Prototype using the IoT Argon Kit integrated with ultra sonic sensors

Prototype using the IoT Argon Kit integrated with ultra sonic sensors

Prototype using the IoT Argon Kit integrated with ultra sonic sensors

We built an advanced prototype using the IoT Argon Kit integrated with ultra sonic sensors, and coded a ruled based algorithm for tracking the bin, shelf capacity and the bin quantity. By doing so we were able to validated balance between technical feasibility and viability of the solution. The following video is a demo to check the bin quantity.

Overview of New Inventory managing & Tracking Experience

ITrack: Mobile app main screens

Research

Field visit to learn user pain points and operations within the AEP department

Field visit to learn user pain points and operations within the AEP department

To identify focus points and develop a feature roadmap, we interviewed the stakeholders and cleared out initial assumptions. Desk Research — We conducted initial desk research to understand the facility’s work. Facility Visit & Interview — We conducted a facility tour and an interview with key management staff. We also conducted an informal competitive analysis of existing software solutions for inventory tracking.

Current manual - visual ways of inventory tracking at the RR dept Toyota manages it inventory

4 Target User roles, their behaviours and workflows

4 Target User roles, their behaviours and workflows

We learned more about them by creating a persona and empathy map. Based on the interview, we sketched out all workflows that are performed by the "AEP Welding Staff” and the vendor.

Overall process of non-scheduled parts management to fulfill PO

Key Findings

Managing non-scheduled parts
currently looked like ...

Managing non-scheduled parts
currently looked like ...

Managing non-scheduled parts looks like ...

Manual effort to track and manage

Manual effort to track and manage

No protocols and Standardized systems

No protocols and Standardized systems

In-person coordination across siloed Departments

In-person coordination across siloed Departments

Unpredictable Lead times and order turn overs

Unpredictable Lead times and order turn overs

Unaware of requirement pattern from clients

Unaware of requirement pattern from clients

Ideation

Converging ideas with the help of value props

Converging ideas with the help of value props

Based on the above key findings and personas, we crafted a Value Proposition Canvas from each persona’s customer segment to guide our brainstorming process; With our holistic understanding of the product, service and the business, we integrated the features with the product architecture. This allowed users to operate with a clear mental model, while navigating the complex interface.

Value proposition mapped from each customer segment

Goals

Improving Experience with defined objectives

Improving Experience with defined objectives

With one mission in mind, we defined objectives to set our product aspirations. This grounded the team's decision making, and build a consensus within the stakeholder group.

Automate

Replace manual effort & reduce human errors by standards for On-the-Go Inventory Tracking.

Simple

Help users easily complete their tasks - nozzle manufacturing rather on Miscellaneous tasks.

Flexible

PO Cycles should adapt to variable levels of inventory information density across verticals.

  • Automate

    Replace manual effort & reduce human errors by standards for On-the-Go Inventory Tracking.

  • Simple

    Help users easily complete their tasks - nozzle manufacturing rather on Miscellaneous tasks.

  • Flexible

    PO Cycles should adapt to variable levels of inventory information density across verticals.

Design

Wire-framing the Interface

Wire-framing the Interface

To get past the analysis phase, we sketched then wire-framed our initial understanding of the feature set. This positioned us to show up with materialized ideas, diverge feedback and converge details on iteration.

Evaluation

Stakeholder’s Feedback

Stakeholder’s Feedback

These were two of our initial high-fidelity prototypes. We conducted a basic heuristic analysis on these and got feedback from our stakeholders. Some drawbacks of these prototypes were that they: 1. Displayed poor user experience — for example, it was tough for users to reach the top dropdown with their thumb. They would have to use two hands in this scenario. Based on this feedback, we improved the next version of the prototypes.

Information architecture of the solution for cross-platform application

Information architecture of the solution for cross-platform application

Solution

Shelf Design and the Infrastructure

Shelf Design and the Infrastructure

The non-scheduled parts are placed in the different bins. The proximity sensors are placed on top of each bin. These help us keep track of whether the bin is present or not, and subsequently help us determine the quantity of the non-scheduled part in that bin -empty, half or full.

Fosters current user pattern of working

The system is fabricated in way that accommodates the nomads such as welders by visually alerting them

Less burden on the managing the bin and more focus on finishing the production orders

Easy to implement and get started with the tracking

Proposed workflow with alerting systems integrated on the shelves

ITrack use cases to address pain-points

ITrack use cases to address pain-points

DESIGN GOAL: SIMPLE

Tedious bin-on-a-shelf tracking

Tedious bin-on-a-shelf tracking

USER SCENARIO

Manual-visual checking of inventory at the bin-level was epicentre to managing the stock and finishing the PO. Since it is Kanban system, empty bins needed to be moved to the top most shelf, later stocked at dock to return to the vendor. Bins goes missing on the shelves, welders sometimes carry them to workflow and forget to put it back.

SOLUTION

We introduced a new shelf design assisted with IoT and ultrasonic detectors, these give a live, informative 360° glimpse of the inventory, status of bin-levels and on-self vs missing bins. The Live tracking gives auto updates and can be refreshed if need be.

DESIGN GOAL: SIMPLE

Their “Go-to + Reliable management” technique: Kanban

Their “Go-to + Reliable management” technique: Kanban

USER SCENARIO

The Dept has been accustomed to use Kanban 2-bin system for their welders and in sync with the Pattonair Ltd (return of the replaced/empty bins to refill). They are relied on this sole system to manage and duly dispatch the week's requirement from Rolls Royce.

SOLUTION

With configuration of digital self to reflect their reality, users can quickly customize to the exact view of the self unit they need. The design accommodates the current signifiers “traffic light” indication to support their mental models.

DESIGN GOAL: Flexible

Selling of non-scheduled parts resulting in reordering

Selling of non-scheduled parts resulting in reordering

USER SCENARIO

When the engines and kits order is fulfilled based on the PO, there cases of surplus of non-scheduled parts. To accommodate space for the inventory of the new PO, some of the non-inventory parts are sold to other vendors.

SOLUTION

Based on the production orders upcoming or current cycles, with AI suggestions user can avoid selling.

DESIGN GOAL: Automate

Communication gaps with complex workflow

Communication gaps with complex workflow

USER SCENARIO

The siloed nature of the workflows and departments, the inventory received goes unnoticed. In turn, the MRPC places an order due to time constraints & unable to reach the dock head for confirmation.
MRPC no or minimal communications with Vendors. Only way MRPC gets to know of order delivery is through email. There are multiple shift manager for the each department.

*MRPC -> Material Requirement Planning Controller

SOLUTION

We redesigned the communication flows by requesting the vendors at each stage of shipment. Within this centralized interface the option to toggle between departments and resend a notification after/during the order delivery process, it enables user to focus on the PO rather than wait on confirmation. The Welders get notified on the commencements.

DESIGN GOAL: FLEXIBLE

Back-order leading to large wait times

Back-order leading to large wait times

USER SCENARIO

Improper bin management -> Improper inventory tracking -> Delay in placing the order. The requirement (quantity and type) of the parts are sporadic and depends on the PO. Third party company was unable to guarantee timely supply:

Request is more than allocated when they have limited stock.

Rise of requests in an unfavourable times.

SOLUTION

Based on the production orders upcoming or current cycles, with AI suggestions user can alert on requirements vendors. once they get confirmation from them they place an order with them

Learning

Delivered Product

Delivered Product

It’s a huge privilege to get ownership of a project where I was able to take advantage of my past experience working at a manufacturing unit as this is one of my products I’ve have delivered with end-to-end solution.

Do more of…

Do more of…

Implementing design system with various component states

We would have iterated more component states to explore the digital real estate and further expanded the AI integration for improved suggestions.

Ways to in-corporate the PO into the system to track at a granular level - the usage of parts.

For our vision based solution, another possible approach would be using image processing using open cv technology. This would have helped us get us real time data with parameter (frequency of parts usage, set different threshold values for each part and etc). In turn, our data analytics would have been more accurate to determine timely availability of these parts and avoid any order issues.

That's me giving demo and briefing on the next steps to the clients

Thank you!

Thank you!

If you have any more questions or want to know more details, please don't hesitate to contact me. For now, please consider checking my other work, my experiments, or learn more about me.