CASE STUDY // 002

Crafting an Experience That Made Drone Autonomy Trustworthy.

BACKBONE OF AUTONOMOUS OPERATIONS

Operator confidence is restored in drones & dock's autonomous system. A fleet operations platform reimagined for the people who depend on it most globally.

COMPANY

INDUSTRY

Robotics, AI

TAGS

Product Design, Osiflow HMI

01 CONTEXT

Autonomy trust doesn't fail in the algorithm. It fails in the interface.

FlytBase powers autonomous drone fleet operations for enterprises globally. Remote missions. Automated flight paths. Multi-docks' coordination across sites.
‍
The autonomy stack was mature. But the people operating these systems were not the people who built them. Security coordinators running night shifts at oil refineries. Site managers reviewing patrol footage across three facilities. Operations leads scheduling daily surveys without a drone pilot on staff.
‍
They needed answers in seconds. Not engineering context in minutes. The interface had to earn their trust. That became the design problem.

02 HIGHLIGHTS

Challenge

Operators running autonomous drone fleets across enterprise sites were making decisions through an interface built by the engineers who built the autonomy. Not designed for them. Confidence gaps slowed every new deployment.

Solution

We restructured the entire operator experience around one principle: every screen must answer three questions in under 5 seconds, covering context, analysis & action.
‍
What's happening now?
Is anything wrong?
What do I need to do?

Outcome

The platform became the global operational standard for drone-agnostic platforms. It was later adopted by DJI - Flight Hub. Serving across oil & gas, construction, security, and infrastructure operations globally.

65%

Enterprise drone adoption is concentrated in mapping, inspection, and agriculture. The operator is rarely the engineer.

V-MR / Persistence Market Research

40%

reduction in operational man-hours when drone-led inspections replace manual methods in the energy sector.

V-MR Industry Report 2026

03 WHERE IT BROKE

The dashboard grew with the engineer. Not with the operator.

FlytBase started as a platform. The first users were drone engineers configuring autonomous flight paths, testing fail-safe logic, and debugging telemetry feeds. The dashboard was built for them. And for them, it worked.

But as FlytBase moved into enterprise industries like oil refineries, construction sites, solar farms, and perimeter security, the person sitting in front of the dashboard changed.

It was no longer the engineer who built the stack. It was a security coordinator on a night shift. A site manager reviewing patrol footage. An operations lead managing drone schedules across three facilities.

They didn't need to understand MAVLink protocols or RTK satellite counts. They needed to know: is my perimeter clear? Did last night's patrol complete? Why is Dock 3 offline?

Operators called engineering to interpret status codes. Alerts were ignored because everything looked the same level of severity. New operators took weeks to onboard instead of days.

The problem wasn't technical. It was experiential.

04 ai approach

Osiflow HMI: An internal AI-native approach to problem solving

Using an internal product used to analyse robots, states, etc. Using the help of Avaline, our AI agent, design & develop outcomes came faster and better than any other agency in the market.

Agentic AI for Human-Machine Interaction design

Avaline understands operator cognitive load and generates production-ready interfaces autonomously.

Production-ready components

Telemetry. Alerts. 3D mapping. Mission timelines. Fleet state. Built once. Deployed everywhere.

Understands the entire system

Every design decision starts from the protocol layer. ROS2. MAVLink. Custom integrations.

Concept to production. In days.

What takes design teams months, Osiflow HMI ships in days. Stratos was built in 3 days.

05 solution

Designing for the person who didn't build it. But has to trust it.

The redesign was a reorientation of who the interface served. Every module was re-evaluated against one question: does this serve the operator's decision, or the engineer's curiosity?

Situational awareness, not information density.

The original command center showed everything simultaneously. Satellite counts, latency metrics, firmware versions, battery curves. For an engineer, this is context. For an operator on a night shift monitoring six drones, this is noise.

We restructured it around the operator's decision flow. Information is always available. It's never in the way.

The alarm system (AMERS) as the trust layer.

Every alert looked the same. Low battery. Geofence breach. Communication loss. Same visual weight. Operators either ignored everything or panicked at everything.

We introduced severity-based hierarchies. Contextual alerts tied to specific devices. Third-party sensor integration with automated responses. The system suggests. The human decides. Trust is built in that gap.

Accountability, not debugging.

Operators don't debug. They answer questions. What happened on Tuesday's 3pm patrol? Why did the drone abort? Can we prove every inspection was completed this month?

We redesigned flight logs as an operator-readable accountability system. Filterable. Replayable. Exportable. Built for operational confidence.

Operational intent, not technical configuration.

Programming a mission meant understanding waypoints, altitude constraints, and geofence logic. Reasonable for a pilot. Not for a facility manager who needs a daily 6AM perimeter scan.

We restructured it around the operator's decision flow. Information is always available. It's never in the way.

What works for one breaks at fleet scale.

Managing fifty drones & docks across ten sites is a fundamentally different cognitive challenge. Not because of the technology. Because the human's attention is finite.

The interface that scales with the fleet, not against the person managing it.

06 OUTCOMES

Same autonomy. Different experience entirely.

The redesigned platform became the operational standard for FlytBase enterprise deployments. Later adopted by DJI.

Adoption

Interfaces that prevent catastrophic operator error and help with quick analysis.

Scale

Few drones view to a complete multi-site fleet management accross teams.

Operator Shift

Engineering & deep-training dependent workflows to self-sufficient workflows for operators.

07 CONCLUSION & testimonial

Improving Customer Experience across the globe in robotics since 2020

With ~4 years of leading & crafting FlytBase & several other products, this case study was to showcase Osiflow's ability to solve your problems.

DRONE fleet AUTONOMY PLATFORM

"I have had the privilege of working closely with Shreshth on multiple projects. He possesses a remarkable ability to approach challenges with a user-centric focus, consistently applying Product, UX & design principles to create innovative solutions.

What truly sets him apart is his dedication to understanding our customers, often stepping into their shoes to effectively address their needs. I observed this commitment firsthand when we collaborated on various custom software integrations, as well as when I sought his insights for technical documents, release notes, and product webinars."

Neel Sharma

Head of Solutions Engineering

Your story could be next.

Prefer Async mode?

Email Us

response time <24 hours avg