What Would the World Miss If T2D2 Didn't Exist?

Facility managers, engineers, and architects across commercial inspection verticals have become inundated with data as drone adoption has accelerated. Capturing drone footage of a large asset was only half the battle. The other half, processing and analyzing potentially hundreds of thousands of images, remained a largely manual, time-consuming task. T2D2 was founded with this growing problem in mind.

Jon Ehrlich, Chief Executive Officer (CEO) at T2D2, a serial entrepreneur with 15 years of experience and two previous companies already under his belt, had always gravitated toward problems rooted in the built environment. Inspections were a natural fit. Engineers, architects, lenders, banks, and governments all require regular inspections for compliance, safety, and asset management. When drone technology began changing how those inspections were conducted, he saw both a gap and an opportunity and T2D2 was born in 2021.

This possibility was the use of machine learning and computer vision for large-scale drone data post-processing. T2D2 was designed to provide inspection specialists with an AI-powered software package that could automatically identify issues across massive amounts of drone imagery, cutting down on the amount of time that would otherwise require months of manual review to a few hours.

The Problem That Would Have Persisted

Without T2D2, the drone inspection industry would have continued to solve one problem while inadvertently creating another. Drones addressed the limitations of traditional inspections, reducing the need for incomplete surveys, repeat visits, and slow manual assessments. But as drone adoption grew, so did the volume of data generated on every job site.

“What a lot of people don't realize is that you can't only adopt drone technology, you also have to pair that with the appropriate workflows and software to provide you with analysis and minimize data overload.”, Jon said.

The expanding use of drones in building inspection, industrial operations, transportation infrastructure, and other sectors is driving greater demand for intelligent post-processing software. It is a structural challenge that will intensify as more industries embrace drones for commercial inspection purposes.

Building Something Engineers Would Actually Use

T2D2 Orthomosaics

One of the more telling insights from Jon concerns how T2D2 approached product design. Rather than building software that would require engineers and architects to overhaul their existing workflows, the team deliberately designed T2D2 to integrate with how professionals already work.

Adoption of new software can stall if it asks too much from the users right away, and professionals in the drone inspection space have limited bandwidth to learn systems that disrupt familiar processes. T2D2 responded by shaping their product around existing engineering analysis practices, making adoption feel like an upgrade rather than a reinvention.

This philosophy extended to data exports as well. After talking extensively with users about their downstream needs, the team built out multiple export pathways, giving users the flexibility to keep data in the cloud, pull it through an API, export in various file formats, or receive tabulated numerical summaries.

The Empire State Building and What It Proves

Perhaps the most striking demonstration of what T2D2 makes possible is its work with the Empire State Building, one of the most iconic structures in the world and a building approaching its 100th year. The Empire State Building engaged T2D2 for a full drone-based inspection of the entire structure from top to bottom.

A local drone service provider captured the data, which was then uploaded to the T2D2 cloud. The project generated over 50,000 photos of the asset. The team developed multiple levels of photogrammetry as well as scan-to-BIM functionality. The drone capture component, which would have taken many months through traditional inspection methods, was completed in a matter of days. Once the data was in the T2D2 platform, computer vision AI condition detection was applied, and deliverables were turned around within hours. The results showed both areas of well-maintained structural condition and areas requiring attention.

This project illustrates the forward-thinking approach of asset owners who understand that preserving aging infrastructure requires embracing better tools.

Setting the Standard

T2D2 has developed its own data capture standard, offering detailed guidance on optimal drone data collection for everything from individual image analysis to photogrammetry and AI processing. The standard addresses variables like lighting conditions, ground sampling distance, hardware selection, camera sensor size, and megapixel resolution.

The T2D2 standard is available for free on the company's website. Everyone's work, including T2D2's own analysis outputs, benefits from improved data collecting throughout the sector. In order to create more comprehensive drone inspection standards, the company has also worked with foreign standards organizations, such as Singapore Standards and the American Society for Testing and Materials (ASTM) in the US.

On the platform itself, automated quality checks assess uploaded imagery before processing begins. If data does not meet the qualifications needed for a given task, users receive an automated notification. The company's support team is available around the clock to help users understand how to improve their data capture before they go back out into the field.

The Core of the Software

At the heart of T2D2 is its computer vision condition detection capability, which the team has spent years refining. The software can automatically detect over 50 types of conditions found in the built environment, including cracks in concrete, spalling, corrosion in metals, and peeling paint, making it the platform's primary value proposition. Beyond condition detection, T2D2 offers in-house photogrammetry tools that generate orthomosaics in both horizontal and vertical planes, allowing users to visualize captured data across multiple dimensions, which is particularly valuable for building envelope assessment. The platform also packs a report generation feature that the team believes deserves more recognition, capable of distilling hundreds of thousands of images down to the most significant findings and packaging them into exportable reports across various formats and templates, though many users tend to stay within the platform rather than exporting finished deliverables downstream.

Looking ahead, the team sees significant opportunity in thermal and infrared imagery analysis. While some platforms offer basic thermal image processing capabilities, and established thermal specialists provide desktop and cloud-based tools, T2D2 views advanced cloud-based thermal analysis as a frontier worth pursuing. The company aims to develop functionality that certified thermographers expect, particularly for engineering-focused applications, and plans to bring this capability to users in the future.

Dispelling the Myth

T2D2 Infrastructure

Jon believes AI in inspection is not a replacement for human expertise, and suggesting otherwise does a disservice to everyone in the industry. The technology is designed to operate somewhere in the range of 70 to 80 percent of human capability on a per-image basis. Where it dramatically outperforms humans is in volume. Processing hundreds of thousands of images in seconds is simply beyond what any human team can do efficiently.

But root cause analysis, structural health determination, maintenance history, and repair methodology all require human judgment. These are decisions that draw on context, experience, and reasoning that goes well beyond visual pattern recognition. AI is improving at incorporating that kind of contextual information as more data is provided about a given asset, but the myth that it will eventually replace engineers altogether is one T2D2 actively works to counter.

The model T2D2 champions is a copilot model, where AI handles the volume and the tedium, and humans handle the interpretation and the decisions that matter most.

Advice for Building in the Drone Space

For entrepreneurs entering the drone inspection industry, Jon emphasizes the critical importance of vertical focus. The commercial inspection sector encompasses numerous applications such as buildings, industrial facilities, transportation assets, telecommunications infrastructure, oil and gas operations, solar fields, wind farms, and pipelines, each with distinct requirements and opportunities.

His advice to newcomers is to identify the specific verticals you aim to serve, and resist the temptation to pursue all of them simultaneously. Instead, focus on a smaller subsegment and develop a compelling value proposition for operators in that space. This concentrated approach allows companies to build deep expertise and establish strong market positions before expanding into adjacent verticals. The drone space offers infinite possibilities when software capabilities are effectively combined with drone-captured data, but success requires disciplined prioritization and a clear understanding of the specific problems you're solving within your chosen market.

What the World Would Miss

Asked directly what the world would lose without T2D2, Jon says, “Well, I think they would miss the T2D2 software, I think they would miss advanced AI condition detection and they would miss things like the drone standard that we've brought to the industry and the role that we've been honored to play over the past five plus years in being a part of that conversation around how AI and in specific, computer vision can augment inspection processes across various industry verticals.”

For a world with aging infrastructure and accelerating drone adoption, that contribution is not a small one.