Innosphere’s NSF ASCEND engine names accelerator cohorts

FORT COLLINS — Innosphere, the Fort Collins-based nonprofit venture development organization that leads the National Science Foundation’s ASCEND Engine, has announced the startups selected for its ARID and SHIELD accelerator cohorts

The 16 companies announced Friday will join two sector-specific accelerators aligned with the NSF ASCEND Engine’s flagship programs: ARID (Asset Resilience through Intelligent Digital Twins), focused on wildfire risk and grid resilience, and SHIELD (Soil Health Innovation, Evaluation and Demonstration), focused on soil health, adaptive grazing, and watershed and water-quality management.

The cohort announcement comes on the heels of national recognition for Innosphere’s accelerator program. Earlier this month, Time named Innosphere to its 2026 list of America’s Best Incubators & Accelerators. The ranking evaluated organizations on founder-reported program experience, portfolio-company track record and endorsements from investors and entrepreneurs nationwide, placing Innosphere among the top accelerators in the country.

It also comes as the ASCEND Engine is taking steps to address environmental intelligence and resilience with three new projects focused on advancing innovation in extreme weather prediction.

Of the 16 companies selected, 10 will be part of ARID and six will be in SHIELD. The six-month, cohort-based programs are designed for companies at Technology Readiness Level 6 or higher, with access to real-world Colorado and Wyoming testbeds.

Cash prizes totaling $150,000 are available per cohort, with $80,000 for first place, $50,000 for second and $20,000 for third place.

“Our ARID and SHIELD cohorts companies are addressing real problems facing our region and country today,” Bud Michael, client director for Innosphere, said in a prepared statement. “These 16 companies are building the tools we need to manage wildfire risk and protect soil and water resources. The recognition from Time is a reminder that the commercialization engine behind them is one of the best in the country.”

Members of the ARID cohort:

• Aftermath Research, which is developing AI-driven, satellite-based damage-mapping technology — including synthetic aperture radar that sees through smoke — to give first responders, survivors, insurers and recovery planners faster, more accurate information after a wildfire.
• Corriente Environmental Solutions (RecordFLOW), which builds low-cost hardware and software to map, measure and monitor open-channel water systems from creeks to irrigation canals, bringing real-time data to utility and agency dashboards.
• DecisionAI, which applies behavioral science-grounded AI to energy systems, shaping when, where and how people use energy to create value for individuals and the broader energy ecosystem, from grids and commercial loads to data centers and other energy-intensive infrastructure.
• Deepfire, which detects wildfires earlier by clustering data from 14 satellites with AI models that filter false positives and route alerts straight to fire agencies and utilities in the field. Its detections have beaten the first 911 call by up to two hours.
• Geofire Labs, which uses satellite imagery and AI to reduce the cost of identifying vegetation that can spark wildfires or power outages along utility powerlines.
• Nostera, which builds roadside risk-intelligence sensor nodes that screen vehicles at highway chokepoints for pre-ignition hazards such as dragging chains, overheated brakes, tire and wheel failures — alerting agencies before a spark becomes a wildfire.
• Omnigeo, a geospatial AI company that measures the physical world from standard imagery, detecting objects, materials and changes in condition across the built and natural environment to deliver property-level wildfire risk intelligence for insurers, utilities  and municipalities. 
• Satellites on Fire, which detects and monitors wildfires using proprietary AI models across a network of more than 13 satellites with a five-minute refresh rate, alerting fire crews and landowners by SMS, WhatsApp, email, or API across 21 countries.
• ThreeV Technologies, which delivers audit-defensible asset condition intelligence for utilities, using multi-sensor vehicles and computer vision to detect wildfire risk and reliability issues at scale, with senior linemen reviewing every major finding.
• WatchPost, which installs purpose-built micro weather stations directly onto utility infrastructure, combining hyperlocal, ground-truth conditions with a utility’s operational data to guide crew deployment and severe-weather response.

Members of the SHIELD cohort: 

• Agroptics, which optimizes irrigation water and energy use through real-time remote sensing and weather forecasting, reducing water consumption by up to 15% while enhancing crop yield.
• BioSensor Solutions — Makes soil biology measurable in real time with a low-cost, printed sensor that continuously reads biological activity below the surface, paired with an AI-driven soil health forecasting platform.
• BioStamp, which synthesizes DNA barcodes for engineered microorganisms, enabling biological intellectual property protection and high-fidelity identification with reduced false positives.
• E-Flux, which provides applied environmental measurement services, supporting research customers with soil and gas-flux data collection.
• Nave Analytics, which fuses satellite data, meteorology and proprietary hydrologic modeling to deliver root-zone soil moisture, crop water demand and drought-risk intelligence without expensive in-field hardware.
• PAGE Technologies, which digitizes water and land systems with low-cost sensors and AI analytics, delivering real-time chemical intelligence for water treatment, agriculture, mining and industrial operations.

NSF ASCEND’s third round of research and translation projects, announced Aug. 11, builds on work initiated during the first two years of the program. The competitive process called on interdisciplinary, collaborative teams from across the region’s leading universities, startup companies and technology innovators to propose decision-support solutions that help communities prepare for and respond to natural hazards.  

The projects selected for funding include: 

• Airborne Intelligent Robotic System for Environmental Sensing (AIR-SENSE), a University of Colorado Boulder program focused on autonomous, fixed-wing navigation to drive precision data collection during extreme weather events.
• AI-Enhanced Hail Forecasting and Ground Validation for Colorado-Wyoming Infrastructure Resilience, a Colorado State University program advancing deep-learning-based hail prediction to support affordability and infrastructure resilience.
• Rocky Mountain Weather and Wildfire Intelligence Center, a University of Wyoming program leveraging the ASCEND-funded RM1.3 dataset to improve spatial resolution of smoke models in complex terrain, with market applications including outdoor recreation and tourism, the health sector and the insurance industry.

“A commonality across projects within our third seedling round is that each team is helping to enable precision interventions that can save lives and protect assets from loss,” Sam Mallory, director of research and development for the NSF ASCEND Engine, said in a prepared statement. “These teams are asking ambitious questions: How can those with respiratory conditions confidently make decisions during heavy wildfire smoke days? Can we help critical infrastructure operators make decisions about moving equipment in advance of extreme hail events?”

These three awards mark the latest expansion of the NSF ASCEND Engine’s growing portfolio of drought, wildfire and extreme weather resilience work and build on the foundation established through the Engine’s flagship ARID and SHIELD programs.

The teams will join the ASCEND Engine’s broader network of researchers, startups and industry partners at the Mountain West AI & HPC Summit, which runs Monday through Wednesday in Laramie, Wyoming. The summit will feature keynote presentations, hands-on training and technical sessions that will include a thematic focus on weather prediction.

More information is on the Mountain West AI & HPC Summit website.

Share on Facebook on LinkedIn