HARVIS · Project

AI Co-Pilot Assistant That Helps Single Pilots Make Better Decisions Under Pressure

transportTestedTRL 4

Imagine you're flying an aircraft alone, and suddenly multiple things go wrong at once — alarms, weather changes, system warnings. HARVIS built a smart digital assistant that acts like a second brain in the cockpit, helping a single pilot handle complex situations that normally require a crew of two. Think of it like an experienced co-pilot who never gets tired, processes information faster than any human, and always keeps calm. The team also created a roadmap telling European aviation where to invest research money over the next decade to make this a reality.

By the numbers

EUR 826,500

EU contribution for developing AI co-pilot technology

4 partners

Consortium size across aviation and AI expertise

3 countries

Countries involved (Spain, France, Italy)

2 demonstrators

Working prototypes built and tested with preliminary results

9 deliverables

Total project outputs including roadmap and demonstrators

50%

Industry ratio in the consortium

The business problem

What needed solving

Airlines and aircraft operators face mounting pressure to reduce crew costs while maintaining safety. Training and retaining qualified pilots is increasingly expensive, and the aviation industry is exploring single-pilot operations as a path forward — but a lone pilot simply cannot process all the information that two crew members handle today. There is a growing need for intelligent digital assistants that can fill that gap without compromising safety.

The solution

What was built

The team built 2 working demonstrators of a cognitive computing digital assistant designed to support single-pilot decision-making in complex flight situations. Alongside the prototypes, they produced a strategic roadmap advising EU aviation research priorities for the next decade across 9 total deliverables.

Audience

Who needs this

Aircraft manufacturers exploring single-pilot certified aircraft (e.g., Airbus, Boeing, Embraer)Avionics system integrators developing next-generation cockpit technologyDefense contractors working on optionally-piloted military aircraftFlight simulation and training companies preparing for AI-assisted cockpitsRegional airlines and cargo operators looking to reduce crew requirements

Business applications

Who can put this to work

Aviation & Aerospace

enterprise

Target: Aircraft manufacturers and avionics system integrators

If you are an aircraft manufacturer exploring single-pilot operations to reduce crew costs — this project developed 2 demonstrators of cognitive computing algorithms that support pilot decision-making in complex situations. The tested prototypes show how a digital assistant can process multiple data streams and recommend actions, potentially enabling certified single-pilot operations for commercial aircraft.

Defense & Military Aviation

enterprise

Target: Defense contractors developing unmanned or optionally-piloted aircraft

If you are a defense company working on reducing cockpit crew requirements — this project built and tested AI-powered decision support that was specifically designed for high-pressure, complex flight scenarios. With 4 partners across 3 countries contributing expertise, the algorithms could be adapted for military transport or reconnaissance aircraft where crew reduction is a strategic priority.

Aviation Training & Simulation

mid-size

Target: Flight simulator companies and pilot training organizations

If you are a training provider looking to prepare pilots for AI-assisted cockpits — this project produced a detailed roadmap for cognitive computing in aviation plus working demonstrators that show how human-AI interaction works in real flight scenarios. The algorithms and use case designs from 9 deliverables could be integrated into next-generation flight training simulators.

Frequently asked

Quick answers

What would it cost to license or adapt this AI co-pilot technology?

The project operated on an EU contribution of EUR 826,500 across 4 partners over 3 years. Licensing terms would need to be negotiated with the coordinator SKYLIFE ENGINEERING SL in Spain. As a Clean Sky 2 project, IP arrangements follow the Joint Undertaking's rules.

Can this scale to commercial aircraft fleets?

The project produced 2 demonstrators with preliminary test results, meaning the technology has been validated in controlled conditions but not yet in operational aircraft. Scaling to commercial fleets would require further certification work and integration with existing avionics systems. The roadmap deliverable specifically maps out what research and development steps are needed next.

What is the IP situation — can we license the algorithms?

HARVIS was funded under Clean Sky 2 (CS2-RIA), which means IP is governed by Clean Sky 2 Joint Undertaking rules. The coordinator SKYLIFE ENGINEERING SL, an SME based in Spain, would be the primary contact for licensing discussions. With 2 industry and 2 university partners, IP may be shared across the consortium.

How close is this to being flight-certified?

Based on the 2 demonstrators with preliminary tests completed by project end in 2021, the technology is at an early validation stage. The project itself produced a roadmap for future development, indicating that significant certification and regulatory work remains before operational deployment.

How does this compare to existing autopilot or crew alerting systems?

Unlike traditional autopilots that follow programmed routes, HARVIS focused on cognitive computing — AI that understands complex situations and supports human decision-making. The digital assistant was designed to help a single pilot handle scenarios that normally require two crew members, going beyond simple automation to active reasoning and recommendation.

What was the project timeline and is the team still active?

HARVIS ran from 2019 to 2021 and is now closed. The consortium of 4 partners across Spain, France, and Italy delivered 9 deliverables including 2 working demonstrators. The coordinator SKYLIFE ENGINEERING is an active aviation SME that may continue development independently or through follow-on projects.

Can the AI assistant technology be adapted for other transport modes?

The cognitive computing algorithms were designed specifically for aircraft cockpits and single-pilot decision support. However, the core approach — AI helping a human operator manage complex situations — could potentially be adapted for maritime bridge operations, rail control, or autonomous vehicle supervision. Based on available project data, no cross-sector adaptation was tested.

Consortium

Who built it

The HARVIS consortium is compact but well-balanced: 4 partners from 3 countries (Spain, France, Italy) with a 50-50 split between industry and academia. The coordinator, SKYLIFE ENGINEERING SL, is a Spanish SME — meaning the technology was developed with a commercial mindset from day one, not just as an academic exercise. With 2 SMEs and 2 universities, the team combines agile engineering capability with deep research expertise. The 50% industry ratio is a good signal for business readiness, as industry partners typically push for practical, implementable results rather than purely theoretical outputs.

SKYLIFE ENGINEERING SLCoordinator · ES
DEEP BLUE SRLparticipant · IT
UNIVERSITAT POLITECNICA DE VALENCIAparticipant · ES
ECOLE NATIONALE DE L AVIATION CIVILEparticipant · FR

How to reach the team

SKYLIFE ENGINEERING SL is an aviation SME based in Spain — reach out to their business development or R&D leadership for licensing and collaboration discussions.

Order a report on this consortium See SKYLIFE ENGINEERING SL profile →

Next steps

Talk to the team behind this work.

Want an introduction to the HARVIS team? SciTransfer can connect you with the right person at SKYLIFE ENGINEERING and brief you on how this AI co-pilot technology fits your specific aviation challenge.

Order a report on this topic More in Transport & Mobility