Exodigo, which maps subterranean spaces with sensors and AI, lands $29M

In the enterprise, drone adoption continues to climb as the commercial applications come into focus — and regulatory restrictions loosen. Enterprise drone use increased by 58% from 2016 to 2017, with the construction, mining, agriculture, surveying, and real estate sectors leading the way. If the trend holds, McKinsey predicts that the economic impact of the commercial drone market — which was $1 billion in 2017 — will increase to over $31 billion by 2026.

Construction, mining, and utility companies in particular have embraced drones for a range of applications, including photography. A Drone Industry Insights report found that inspection, mapping and surveying, filming, and localization and tracking are among the top reasons that businesses operate drones. While drones aren’t a perfect technology — even the best models have limited range and capacity — respondents to the report said that adopting drones enabled them to cut costs and save time.

Often, companies turn to third-party vendors — e.g., DroneDeploy, DroneBase, and PrecisionHawk — to help orchestrate their drone programs. A newer player in the space is Exodigo, which claims its software can create 3D maps of underground man-made pipes, cables, soil layers, rocks, minerals, and groundwater. Alongside the availability of its platform, the company today announced $29 million in seed funding co-led by Zeev Ventures and 10D Ventures with participation from SquarePeg Capital, JIBE Ventures, Tidhar Construction, Israel Canada, and WXG Ltd.

Drones for inspection

Exodigo was cofounded in 2021 by Jeremy Suard and Ido Gonen. Both served in the Israeli Military Intelligence, where Suard led AI and deep learning teams specializing in signal processing and acoustic engineering. Prior to the military, Gonen was a teaching assistant at Technion, the Israel Institute of Technology, in Haifa.

“We deliver a digital map that shows you what is actually there, without the need for interpretation — and can scan vast areas at scale with high accuracy using drones or more complex urban environments with our high-precision ground platforms,” Suard told VentureBeat via email. “[O]ur breakthrough solution combines fundamentally different signals (like vibrations, magnetics, and electricity) into a single normalized matrix to train the AI platform on.”

According to Suard, 35-employee Exodigo leverages a combination of “non-intrusive” subsurface imaging technologies, fusing data from on-drone or on-cart sensors to detect buried objects “at any scale or terrain.” In its first live demonstration, Exodigo reportedly identified underground utilities, abandoned lines, and ground layers in a remote, unmapped area in Israel.

Subsurface mapping isn’t an easy feat. As a 2021 research paper notes, detecting soil underground tends to be especially challenging because soil absorbs or attenuates most of the electromagnetic spectrum, preventing sensors from identifying all but the surface of the soil. “Even in the case of the same soil, the electromagnetic properties (propagation speed of electromagnetic waves, relative permittivity, etc.) may vary depending on the water content or medium condition,” the coauthors wrote. “In other words, there is a concern that the depth evaluation of underground objects by engineering judgement alone may cause significant errors in detection and estimation of depth.”

Suard says that Exodigo’s use of multiple sensors and “proprietary AI” helps to overcome some of these challenges.

“Our [platform] provides users with a safe, fast way to get a complete view of what lies beneath the surface with unprecedented precision,” Suard said in a statement. “Ending the era of blind digs, Exodigo gives companies an accurate, easy-to-understand map of what lies beneath the surface – empowering their teams to save time, money, and lives. Think of it like combining the scanning power of an MRI, CT scan, and ultrasound all into one image of what is beneath the ground.”

Exodigo — which is pre-revenue, and plans to commence pilot projects in California, Florida, and Texas in the weeks ahead — says it’ll use the funding to support expansion with a focus on building a California-based team.

“Because of our unique access to government support, we have been able to use dozens of terrains with ground truth for AI training across a range of terrain types and a diversity of utilities. This allowed us to generate proprietary data and create an incredibly large and diverse dataset,” Suard continued. “Our core expertise and main advance in AI is how to normalize low signal-to-noise ratio and super noisy signals into matrices that AI is used to train on. We excel at multichannel signal processing and the ‘needle in a haystack’ detection problem … We have successfully completed demonstration projects with a number of design partners such as the Israeli water company, a municipality in North Israel, and top Israeli construction companies.”

Expanding market

Some 26% of small to midsize construction businesses either are — or expect to — incorporate drones into the operations by 2020, according to a 2018 Software Connect survey. Insider Intelligence predicts total global shipments to reach 2.4 million in 2023. And drone use in construction and mining alone could eventually become a $28.3 billion global market, according to PricewaterhouseCoopers.

Exodigo competes with Emesent, which is developing software that enables drones to map subterranean spaces autonomously. Swedish startup Inkonova also claims to have developed autonomous inspection drones for underground mining. Beyond drone-based solutions, companies like 4M Analytics and Pulse Utility map utilities burried deep below ground using an array of sensor technologies.

But Suard asserts that Exodigo’s solution is differentiated by its preciseness and speed.

“Top use cases for our AI platform include industries like construction, mining, utilities, etc., but our future vision is much bigger. Our imaging technology has no boundaries — quite literally,” he said. “We can see applications as close to home as supporting survivor-finding missions when buildings, tunnels, or mines collapse; as deep underground as transportation tunnels for the next era of robotic digging crews; or as far away as supporting space exploration missions like the Mars Rover.”