Autonomous ground vehicles—or “steel scouts”—have long been heralded as the logical extension of unmanned air systems onto land, promising to push the reconnaissance envelope and extend force projection. Yet the U.S. Army’s Robotic Combat Vehicle (RCV) program finds itself at a critical crossroads: despite demonstrators featuring hybrid powertrains, tele-operated payloads, and AI-driven navigational kernels, worries over cost-effectiveness and victories handed to $200 drones over $3 million bots have spurred a top-down reevaluation. In June 2025, Business Insider reported Army leadership rethinking the RCV’s value proposition amid rising attrition risks. Just weeks earlier, Breaking Defense revealed plans to cancel a planned RCV award and pause howitzer competitions in an 8 percent budget realignment, underscoring the program’s fragility. This confluence of promise and peril sets the stage for a defense community in need of modular, low-cost solutions and doctrinal updates to unlock autonomous platforms on future battlefields.

In 2019, the Army launched the RCV program as part of its modernization drive, assigning a triad of classes—Light (RCV-L), Medium (RCV-M), and Heavy (RCV-H)—to mirror the manned fleet’s operational spectrum. Early concepts envisaged robotic scouts teamed with M2 Bradleys, Strykers, and Abrams tanks, offering overwatch or decoy functions while keeping crews au courant with AI-driven situational awareness. By 2023, budget pressures and tech hurdles distilled efforts onto the Light RCV, with RCV-M and RCV-H shelved. The 2025 “Army Transformation and Acquisition Reform” further crystallized this realignment, formally halting RCV development in the face of uncertain operational value and autonomy integration challenges. This shift underscores the Army’s broader pivot away from legacy platforms toward leaner, networked systems optimized for contested environments.

Although the program once spanned three weight classes, prototyping has ultimately orbited around the Light RCV class, yielding four demonstrators: Textron’s Ripsaw M3, General Dynamics’ TRX, QinetiQ-McQ’s WOLF-X, and Oshkosh Defense’s modular entry. Each features distinct powertrains (diesel-electric hybrids to conventional diesels), tele-operated or semi-autonomous weapon stations, and cell-based comms stacks capable of platoon-level C2. Underpinning them all is the Robotic Technology Kernel (RTK), the service’s in-house autonomy software meant to provide a unified toolkit for perception, navigation, and mission management. Industry sources have lamented the RTK’s developmental pace and the Army’s piecemeal payload acquisition, arguing that cobbling autonomy, sensors, and weapons without a clear competition strategy risks delayed fielding and interoperability headaches. Yet proponents counter that field experimentation is the only path to mature capabilities, with early units slated for operational test in select brigade combat teams. Understanding these prototypes’ performance and limitations is key to charting an RCV roadmap that balances ambition with pragmatism.

As the RCV prototypes matured, so did the ledger entries: industry sources pegged unit prices north of $2.5–3.0 million per vehicle, rivaling manned platforms’ acquisition costs but without the proven operational pedigree. This disparity became stark when $200 commercial quadcopters demonstrated they could blind or destroy RCV sensors at a fraction of the price, prompting questions about return on investment. In May 2025, an internal Army memo revealed that leaders paused the RCV award in an 8 percent budget drill, preferring to re-open competition rather than commit to a single $3 million copy. Defense Secretary Pete Hegseth has since pressed for a leaner, more lethal force posture—one that delivers effects at scale while avoiding “sunk cost” traps. The calculus now demands that autonomous ground vehicles not only perform novel missions but also outvalue adversary drones, making cost-per-mission metrics as critical as technical capability.

Autonomous ground vehicles operate in a battlespace that no longer tolerates RF silence: adversaries have fielded electronic warfare (EW) kits capable of jamming GPS, disrupting line-of-sight data links, and spoofing autonomy algorithms. The 2023 Trusted Operation of Robotic Vehicles in a Contested Environment (TORVICE) trial in South Australia under AUKUS highlighted these risks, running U.S. and U.K. ground robots through a gauntlet of EW weapons to evaluate how UGVs react under duress. The trial revealed that while basic obstacle avoidance held firm, compromised comms chains forced fail-safe behavior that shut the vehicles down or sent them to pre-set loiter points. Beyond jamming, RCVs face cybersecurity threats: embedded networks and open-standard protocols expand the attack surface for malicious code injection, supply chain tampering, or data exfiltration. Environmental factors—urban canyons, foliage, and rough terrain—further compound sensor fusion challenges. Bridging these gaps requires a holistic risk mitigation strategy encompassing hardening, redundancy, and rigorous cyber testing before any fielding decision.

To reconcile aspirations for autonomy with fiscal realities, the next iteration of RCVs must embrace modularity and attritability—design paradigms successful in UAS programs but underexplored in UGVs. The Army’s June 2025 push to streamline procurements underscores this shift: Secretary Driscoll emphasized the need to “close the loop on being able to purchase those things more quickly so that … small and medium businesses … can get rewarded for those investments.” One promising blueprint borrows the commercial “kit” approach: off-the-shelf self-driving stacks retrofitted onto modified Infantry Squad Vehicles, as Applied Intuition demonstrated by autonomizing a Chevy Colorado in ten days. Breaking an RCV into a chassis, powertrain, autonomy suite, and mission payload allows rapid COTS upgrades and cost amortization across platforms. Attritable Light RCVs could be fielded en masse, leaving a smaller budgetary footprint per deployed system and enabling sacrificial tactics to saturate contested zones. This composable approach could lower per-unit costs, mitigate advancement risk, and accelerate learning cycles on the ground.

Hardware advances mean little without doctrine that fully integrates manned and unmanned assets. Early RCV concepts imagined “robots on leashes,” but battlefield realities demand flexible comms and revised tactics. Industry has flagged that the Army’s reliance on RTK and legacy line-of-sight radios undercuts operational reach. New mesh networking waveforms—MIMO combined with OFDM—allow signal “bouncing” around urban obstacles, while satellite-augmented C2 links offer resilience beyond line-of-sight. Yet each solution carries trade-offs in latency, bandwidth, and signature management. Shifting from teleoperation-centric missions to beyond-line-of-sight autonomy requires rewriting Multi-Domain Operations doctrine to codify new roles for RCVs: from distant scouts inserting into enemy flanks, to decoy platforms drawing fire, to automated resupply convoys. Doctrine must also clarify ROE, sensor-to-shooter pipelines, and soldier-in-the-loop thresholds to manage human–machine teaming effectively. Only then can steel scouts avoid turning into logistical burdens or collateral liabilities.

To navigate the promise and peril of RCVs, policymakers and program managers should:

• Adopt a modular open systems architecture (MOSA) to enable COTS autonomy and payload swaps.
• Prioritize attritable Light RCV variants for early fielding, leveraging scalable production lines.
• Invest in EW- and cyber-hardened networks—mesh waveforms, satellite backhauls, and encrypted data links.
• Integrate RCVs into updated Multi-Domain Operations doctrine, specifying roles, ROE, and human–machine interface standards.
• Establish a rapid prototyping and experimentation cell within Army Futures Command to iterate on lessons learned.

The RCV saga illustrates the knife-edge between technological promise and fiscal discipline: without clear value propositions, even the most advanced bots risk budgetary extinction. The 8 percent realignment that paused the RCV award exposed a sobering truth: force multipliers must prove cost effectiveness against low-cost adversaries and integrate seamlessly into existing formations. Yet dismissing autonomy wholesale would forfeit the chance to field steel scouts that can scout ahead, screen mechanized formations, and disaggregate high-risk missions. By embracing modular, attritable architectures and embedding RCVs within a refreshed doctrine, the Army can harness unmanned ground vehicles as true force multipliers rather than white elephants. The path forward demands patience, pragmatism, and a willingness to learn from every line-of-sight failure and jammed comm, lest tomorrow’s battlefields favor the cheapest killer drones over tomorrow’s robotic cavaliers.