The U.S. Marine Corps just handed one of its least glamorous but most tactically urgent problems to two completely different teams running completely different autonomy stacks. The problem is simple: keep forward-deployed Marines alive by delivering ammunition, water, and medical supplies without putting a helicopter crew in the line of fire. The solution that emerged was not simple at all, it was a split-award competition that puts a Pittsburgh commercial autonomy startup and a Lockheed Martin subsidiary on parallel prototyping contracts, each racing to prove their autonomous cargo helicopter works better than the other.
The program is MARV-EL, short for Medium Aerial Resupply Vehicle, Expeditionary Logistics, and it has been running in some form since 2023. What changed now is the Marine Corps stopped hedging and started competing. Earlier phases explored concepts. Increment 2, the version just awarded, requires actual flying aircraft, actual payload capacity, actual combat radius, and two teams building toward the same performance threshold simultaneously. The winner gets follow-on production consideration. The loser gets a data point. This matters because split awards signal something Pentagon procurement almost never admits: uncertainty about which technology actually works. When the Pentagon is sure, it writes a specification and picks one vendor. When it is not sure, it funds two teams and watches them prove it in the field.
Sikorsky, the Lockheed Martin subsidiary, received a $15.5 million contract to deliver the R66 TURBINETRUCK, an autonomous cargo version of Robinson Helicopter's R66 airframe outfitted with Sikorsky's MATRIX autonomy system, which has been flying for years and is proven in government testing. The R66 is light, turbocharged, and operationally familiar to military helicopter communities. Near Earth Autonomy, the startup founded by Carnegie Mellon researchers and based in Pittsburgh, received a separate OTA (Other Transaction Agreement) to develop an uncrewed Bell 505. Near Earth pairs its own Captain autonomy stack with Bell's larger, more powerful 505 airframe and Moog's Genesys avionics. Both teams must meet a 1,300-pound payload requirement and 100-nautical-mile combat radius. Near Earth's proposal claims it can fit an entire Joint Modular Intermodal Container inside the 505 and get two aircraft into a C-130 with minimal disassembly, a logistics multiplier the R66 team does not advertise.
Why split it now? Three things converged. First, the 2023 Afghanistan withdrawal exposed the visceral risk of crewed resupply missions in contested terrain, something no amount of prior operational experience actually solved. Second, autonomy stacks matured enough that two competing vendors could each claim credible path to a working system by late 2026 or early 2027. Third, the Pentagon's Other Transaction Authority vehicle matured as a contracting mechanism, letting NAVAIR (Naval Air Systems Command) award Near Earth without the acquisition overhead that would have killed the deal in a traditional Small Business Innovation Research pipeline. The OTA to Near Earth and the direct contract to Sikorsky are the same signal: the government is ready to buy autonomous vertical-lift capability if it works, and it is willing to pay for two parallel prototypes to find out which design actually works at scale.
The competitive dynamics are asymmetric in a way that matters. Sikorsky brings institutional credibility, proven autonomy pedigree, and $15.5 million in funded development. Near Earth brings agility, a commercial autonomy stack already deployed on other platforms, and access to Bell's production expertise, plus the psychological advantage of being the David in this fight. The R66 is lighter and requires less infrastructure. The Bell 505 is bigger and claims better payload efficiency. Neither team is wrong; they are just optimizing for different mission parameters. When the prototypes fly head-to-head in Marine Corps evaluation, likely in late 2026 or early 2027, one autonomy architecture will prove more reliable or more maintainable or more cost-effective per sortie. That winner becomes the production baseline.
Watch three things. First: prototype delivery timelines. Both teams announced target dates around Q4 2026; slip past January 2027 and you will know one team is struggling with autonomy integration or airframe integration. Second: Marine Corps testing results, specifically failure modes. If one autonomy stack shows degraded performance in GPS-denied environments or fails more often in failsafe scenarios, that determines the winner regardless of payload capacity. Third: cost-per-flight-hour estimates as prototypes mature. The Pentagon will eventually compare not absolute capability but cost-to-keep-a-resupply-bird-flying at the tactical edge. If Sikorsky's MATRIX-on-R66 costs 40% less per hour to operate than Near Earth's Captain-on-505, that narrows the competition down to one clear choice, even if the other design is technically superior.