inQ Services

Building confidence in facilities management during troubleshooting

Work in collaboration with inQ Services , a startup specializing in IIoT and sensor management
Timeline
Jan 2023 - Jul 2023
My role
Product Designer
Product Manager
Team
3 Product Designers
1 UX Researcher
Overview
As a part of CMU's MHCI program, this capstone project is a partnership with inQ Servicesa Virginia-based tech startup who is developing a B2B SaaS ptform to manage sensors in the domain of industrial IoT. My team and I have been tasked to increase the value of their product, inQntrol, through a research-based and human-centric user experience design. Aspiring to be universal and adaptable to any industry, the capstone project is tackling higher education as the first case study. I led the effort of product thinking, problem framing, interaction design, visual design, defining success matrices, user testing, as well as project planning.
PROBLEM
The workflow of facilities management currently involves many disparate digital platforms that are mostly text-based, leading to operational inefficiency during troubleshooting.
SOLUTION
An ecosystem of a desktop application and a mobile application that helps cut down troubleshooting time through providing interoperability and spatial context.
IMPACT
  • 11 out of 11 users from 5 institutions affirmed that the product would reduce the onboarding time for new hires
  • Client applied project’s findings to their core product and planned on implementing our product
01

Context

Background Context

Facilities management = maintaining comfort, efficiency, and safety in a college campus

Facilities management is a maintenance of an institution’s physical infrastructure to ensure a smooth operation of its facilities. The term encompasses many tasks, but for the scope of this project, we focused on the regular upkeep and repairs of building systems, such as electrical, plumbing, HVAC, etc.

End Users + Existing Workflows

Facilities manager monitor equipment and assign work orders. Technicians troubleshoot and fix the problems.

A challenging problem space

The existing software do not communicate with one another and mostly rely on textual information.

Bookmarked links on a facilities manager's browser
Most management software in the domain are text-based
02

Research

Understanding Problem Space

Through extensive research, I learned that the most universal pain points in facilities management are alert/platform fatigue and lack of manpower.

Through months of research, I led the team through multiple research methods that unpack this highly complex and technical domain.

4 research insights

INSIGHT #1

Alerts should empower, not overwhelm

Facilities managers experience alert fatigue (feeling desensitized to alerts) due to the overwhelming number of alerts that lack insights and actionable next steps.
INSIGHT #2

Troubleshooting mechanical problems can be a guessing game

Current software lack a holistic view of the different systems in facilities management, making troubleshooting issues more challenging.
INSIGHT #3

Sensor data reveals equipment health

Sensor readings indicate the condition of the equipment they are attached to.
INSIGHT #4

Document now for better analysis later

Facilities management lacks a proper system to document sensor info, resulting in disorganized data over time that prevents accurate predictive analysis.
03

Problem Framing and Ideation

Unpack User Needs

Finding the problems worth solving

I further drilled down the final 4 insights to identify the true user needs.

Unpack user needs on the whiteboard

3 user pain points we are tackling:

1
Alerts are not actionable
2
Software do not communicate with one another
3
Alerts lack spatial information
Ideate

Tested 16 ideas for desirability with users and clients

The team brainstormed ideas for the smart assistant and tested them with inQ’s CEO, product managers, graphic designer, and engineer, as well as 9 potential users through a prioritization activity. Based on this activity, we were able to derive that the top three most “invested” ideas were a notification center, spatial context, and root cause analysis.

Design Principles

Synthesized feedback into 3 design principles

Through the testing sessions, we gathered feedback and synthesized it into the following three key takeaways:

1. Promote Actionability

Group alerts by zones, not criticality, since most alerts come in marked with high criticality

2. Encourage Curiosity

Do not provide one correct answer. Simply provide resources for users so they are equipped with what they need to solve an issue.

3. Harness Spatial Awareness

Be the map. Providing more spatial context will cut down time to search for a room or an equipment and help onboard new personnel more quickly.
Touch Points

3 important touch points

With spatial context in mind, I pushed the team to focus on three main touch points:
when users receive alerts (1), when they make sense of the problems off-site (2), and when they troubleshoot on-site (3).

Two modalities for different touch points: a desktop app for virtual troubleshooting and a mobile app for on-site troubleshooting.

Because of the two different settings where our solution could bring value, I determined that inQ needs to provide an ecosystem of products that serves different nature of the work.

04

Iteration

Iterated on the design through 3 rounds of user testing with 15 facilities managers and technicians across 6 institutions

Learnings and subsequent design updates are summarized below:
Desktop

I learned that users liked the idea of inQntrol as a directory of resources for troubleshooting, but the medium should be more familiar to them

Hoping to integrate into the existing workflow of a facilities manager/technician as seamlessly as possible, I first proposed a widget as an interface for the desktop app. However, through testing, we learned that the widget introduces a foreign mental model and isn't ideal. We also learned that users really liked the idea of consolidating resources and linking them into a single platform like inQntrol.

Earlier iteration of the desktop app as a widget

Based on user testing, I led the team on the following updates:

BEFORE

A widget introduces a new, unfamiliar workflow

Even though it integrates more seamlessly into the existing workflow, a widget introduces a new mental model that's too foreign

AFTER

A standalone app builds on existing mental model

Switching to a standalone platform matches with the existing workflow more and is easier to understand

BEFORE

Overly complex 3D model

Despite how equipment could be at different elevations within a space, an interactive 3D model becomes too excessive and overwhelming

AFTER

Simpler and still informative 2D diagrams

2D diagrams are more straightforward and still comprehensive

BEFORE

Linking only to building automation systems (BAS)

Linking to BAS is useful, but this unique feature could do much more

AFTER

A directory of many resources

inQntrol can act as a directory of many resources that supplement the troubleshooting process, ex. vendor websites for ordering parts for replacement

Mobile App

Users refer to past work orders and technical drawings the most while on site, and AR turned out to be the most anticipated feature

Through user testing, I learned that, consistent with our second design principle, users prefer to have more information at their fingertips that they can reference without being told what actions to take. I learned what information they prioritize more and adjusted the design to fit that hierarchy.

AR was surprisingly a big hit. I decided to tie in the resources to help troubleshoot with this feature.

Initial screens for mobile app

Based on user testing, I led the team on the following updates:

BEFORE

Too many root cause suggestions

Offering 5 potential root causes in fact raised more skepticism around the recommendations

AFTER

Present only top 3 potential causes, along with resources

Displaying the top 3 likely causes, along with resources for further investigation, is more impactful

BEFORE

Disconnection between alert and related past work orders

Users intuitively expected the comprehensive past work order screen to be specific to the most recent work order

AFTER

Display past work orders upfront

We added the recent work section to each individual alert so the flow is more intuitive and convenient

BEFORE

Plans show limited info

Design previously included 2D plans related to the specific department each user belongs to (ex. HVAC)

AFTER

Flexibility to overlay other technical drawings

There's an opportunity to overlay drawings from other departments (ex. plumbing and electrical) to provide a more holistic overview of the space

05

Final Solution

Virtually troubleshoot from your office via the desktop app:

Receive service requests

Get a quick snapshot of all alerts on campus, grouped into zones

Identify cause of problem

Virtually investigate an issue through spatially mapped sensor readings and linked resources at the building level...

Identify cause of problem

...or zoom in closer to look at each floor

Once on site, the mobile app helps you investigate and fix an issue more effectively:

Receive work order

View potential root causes and previous work related to an alert, organized in a map view

Fix problem

Customize the displayed information on a floor plan to quickly locate an equipment

Fix problem

Confidently locate parts and access relevant info on site via AR

Visual Design

A clean look consistent with inQ's branding

Optimizing inQ's unique branding color pink, I created a set of UI components that are clean and align with inQ's image of friendliness and simplicity.

06

Impact

Implementation

Due to the promising impact of our design, inQ is looking to have the design implemented and beta tested.

Success Metrics

Via beta testing, I hope inQntrol would meet the following criteria:

07

Reflection

What would I do differently?