Palladyne AI (PDYN) – 2026 roadmap, dual engines and embodied AI for defence & industry

Palladyne AI Corp. is the new name and structure of the former Sarcos Technology and Robotics Corporation, historically a robotics hardware company with roots going back to the 1980s. After a difficult public-market journey as a hardware-heavy SPAC story, the board executed a radical pivot: exit from capital-intensive robotics hardware and refocus on embodied AI and autonomy software for defence and industrial markets.

In March 2024 Sarcos formally became Palladyne AI, with ticker change from STRC to PDYN. The new brand reflects the narrower scope: less CAPEX-driven robotics integration, more software, autonomy and high-value engineering services. The company now defines itself as “America’s cross-domain force multiplier”, with a focus on AI that can be embedded into physical systems rather than cloud-only analytics.

The core of Palladyne AI is a closed-loop autonomy software portfolio designed for “embodied AI”, i.e. AI running directly on machines (robots, drones, missiles, industrial cells) rather than purely in the cloud. The three main software pillars are clearly laid out in the 8-K slide deck:

2.1 Palladyne IQ (industrial & defence)

Palladyne IQ is the foundational AI platform, originally built for industrial robots and later extended to defence use cases. Version 2.0, announced in early 2025, is a secure, hardware-agnostic autonomy stack for complex robotic tasks:

• Closed-loop autonomy, with ability to sense, plan and act in dynamic, unstructured environments.
• Domain-specific language models and probabilistic ML to generalise from only a few demonstrations.
• Designed to work with most industrial robots sold today, without requiring heavy cloud connectivity.
• Supports “language-to-motion” type instructions, suitable for edge computing and constrained networks.

From a commercial standpoint, IQ aims to create a recurring software and services revenue stream in industrial automation (manufacturing, logistics, infrastructure), while also serving as the AI core for some defence programs.

2.2 SwarmOS and Palladyne Pilot (defence & UAV autonomy)

On the defence side, SwarmOS and Palladyne Pilot extend the IQ architecture to mobile platforms:

• SwarmOS™: closed-loop autonomy for UAV swarms, built to coordinate multiple drones and sensors with game-theory-driven task allocation and reinforcement learning for sensor control.
• Palladyne Pilot: autonomy layer focused on detection, tracking and control of mobile machines (robots, UAVs), controlling sensors but not flight/navigation itself.
• These platforms are tightly integrated with the company’s acquired UAV hardware and BRAIN avionics modules.

2.3 Hardware, components and manufacturing

Unlike typical pure-software AI plays, Palladyne now owns critical hardware and manufacturing assets through the acquisitions of GuideTech, Warnke and MKR in late 2025.

• GuideTech: aerospace design and avionics engineering, proprietary UAVs and flight controllers, able to move from digital model to flight prototype in under six months.
• Warnke & MKR (formerly Crucis): certified US manufacturers supplying major defence programs (F-16, F-35, Tomahawk, Harpoon, Bradley, JDAM, M1A1, etc.).
• Palladyne Manufacturing: integrated machine shops and assembly, bringing total manufacturing and assembly capacity to ~120,000 square feet in the US.

This vertical integration is positioned as a structural advantage: the AI core, UAVs and components, and certified manufacturing live under one roof, which should matter in the context of US DoD reshoring and the Replicator initiative.

The “Palladyne Defense” business aggregates SwarmOS, BRAIN avionics, loitering munitions (Project Banshee), cruise-class systems (Project SwarmStrike) and the Warnke/MKR manufacturing footprint. The pitch is clear: a “new mid-tier prime” that can move faster than large primes but scale more than small startups.

The “Palladyne Commercial” side takes the same autonomy technology and applies it to industrial robots and automation. The IQ 2.0 platform, officially launched in early 2025, is designed to plug into existing robot fleets without requiring full system replacement.

4.1 Use cases and value proposition

• Higher utilisation of installed robots through more flexible task learning and reduced integration work.
• Ability to handle dynamic environments (variable parts, layouts, disturbances) with fewer manual re-programming cycles.
• Potential to unlock “edge autonomy” without depending on expensive, latency-sensitive cloud inference.
• Target sectors: manufacturing, logistics, infrastructure and some infrastructure-adjacent defence uses.

The slide deck explicitly frames Palladyne IQ as delivering a recurring software licensing and services model – a structure that, if it gains traction, could smooth the otherwise lumpy defence-contract revenue profile.

4.2 Status as of early 2026

• Palladyne IQ development started around 2019, with MVP completed in 2024 and initial customer trials starting in late 2024.
• By the time of the Needham deck, the roadmap includes additional STRATFI funding and continued trials into 2025/26.
• Commercial success is still to be proven at scale; 2025 revenues remain small, and the 2026 guidance assumes a mix of acquired revenue (engineering/manufacturing) plus new AI development contracts, not yet large-scale ARR.

The FY 2024 10-K shows the end of the heavy restructuring phase: asset write-downs and large R&D cuts tied to the hardware shutdown are largely behind the company. Operating expenses in 2024 fell sharply compared to 2023 across R&D, G&A and sales & marketing.

The January 14, 2026 8-K (Exhibit 99.1) is the main source for the 2025 preliminary figures and for the 2026 outlook. The key slide summarises revenue, cash, backlog and expected cash burn.

6.1 2025 preliminary snapshot

• Revenue 2025 (prelim): $5.0–5.5M, below the 2024 level, reflecting the pivot and transition.
• 12-month backlog: >$13M as of 12/31/25, significantly higher than one year earlier.
• Cash & marketable securities: roughly $47M at 12/31/25.

6.2 2026 guidance and implied growth

• 2026 revenue guidance: $24–27M (roughly 4–5× versus 2025 prelim).
• Cash burn: ~$6–7M per quarter in 2026, in line with 2025, excluding a planned $5M development spend over 12–18 months to bring Banshee and SwarmStrike to full commercial readiness.
• Backlog: management expects the majority of the >$13M backlog to convert into revenue within 12 months.
• Profitability: acquired businesses (engineering services + components) are targeted to be adjusted EBITDA-positive in 2026, but the group overall is still in investment mode.

The guidance therefore hinges on two main contributors: (1) consolidation of acquired revenues and ramp-up of defence components and manufacturing, and (2) conversion of AI development contracts into deployed systems and more recurring revenue over the 2026–2028 window.

As with most small-cap tech names emerging from a SPAC cycle, capital structure is a critical part of the PDYN story. The 10-K and 10-Q detail several financing steps in 2024–2025:

• October 31, 2024: announcement of ~$7M in gross proceeds from the sale of common stock and warrants via investor and insider purchase agreements.
• November 13, 2024: entry into an open-market sale agreement (ATM) with Jefferies, initially for up to $18M of common stock, later expanded by an additional $30M on December 31, 2024.
• From January 1, 2024 through the 10-K filing date, the company had already sold several million shares under the ATM, raising meaningful proceeds.
• In 2025, the company also filed a prospectus related to the potential resale of shares by existing holders, giving a framework for secondary liquidity.

The net result is a balance sheet that is substantially derisked versus the low cash position of legacy Sarcos, at the cost of higher share count and overhang from registered resale securities. For a run-up-style strategy, the presence of an active ATM and potential selling stockholders is an important part of the risk picture.

The company breaks its roadmap into a simple three-year narrative: 2026 is “crawl”, 2027 is “walk”, 2028+ is “run”. The slide deck outlines the operational focus for each phase.

8.1 2026 – “Crawl”

• Prove, integrate and de-risk the combined AI + UAV + manufacturing platform.
• Execute on development contracts and integration work with partners (including Red Cat and Draganfly for SwarmOS/IntelliSwarm demos).
• Advance Banshee and SwarmStrike toward higher TRLs and near-operational readiness.
• Expand Palladyne IQ trials and pilot deployments in industrial environments.

8.2 2027 – “Walk”

• Conversion from validation to contracts: UAV wins, defence programs and early recurring revenue (ARR) tied to IQ and SwarmOS.
• Improving margins and operating leverage as acquired businesses scale.

8.3 2028+ – “Run”

• Broad deployments across enterprise and government customers, multi-year contracts and potential platform standardisation.
• Optionality from allied country wins, additional M&A and platform extensions.

From a catalyst perspective, 2026 will mainly generate qualitative triggers (demos, trials, incremental contracts) rather than binary events, but any visible programme win in Banshee, SwarmStrike or a large IQ deployment could change market perception quickly.

The leadership team is built around the pivot from hardware robotics to AI and autonomy. Laura Peterson, with prior experience at Boeing, leads the company as CEO, supported by a CTO with 25+ years in AI/ML according to the company’s materials.

• The deck emphasises 30+ years of robotics engineering DNA and over 140 team members at 12/31/25.
• Recent acquisitions added senior missile/UAS design engineers and experienced manufacturing executives with long histories supplying major defence programmes.
• Board and governance remain those of a small-cap tech stock, with all the usual volatility and sensitivity to equity-market conditions.

Strategically, management is clearly aligning with US defence priorities: attritable systems, AI in mission systems (not just analysis), reshoring of manufacturing and “cost-per-effect” as the key metric.

Unlike a biotech with fixed PDUFA dates, PDYN’s catalyst map is a sequence of programme milestones, demos and contract wins. From the slide deck and public information we can extract a non-exhaustive list of areas to watch:

• Palladyne IQ: additional customer trials, live demos and, ideally, named deployments with major industrial players.
• SwarmOS / IntelliSwarm: demos and contract progress with partners such as Red Cat and Draganfly, plus any direct DoD contracts referencing the architecture.
• GuideTech / Banshee / SwarmStrike: development milestones, flight test campaigns, selection in specific programmes or as part of larger systems.
• Palladyne Manufacturing: expansion of capacity utilisation, new contracts with primes and Tier-1s, and cross-selling of AI-enabled manufacturing optimisation (using IQ internally).
• International / allied markets: any early wins in allied countries would validate the scalability of the model beyond core US programmes.

The company’s 10-K and 10-Q include an extensive risk section, but a run-up-oriented investor should keep at least the following points in mind:

• Execution risk: the 4–5× revenue ramp from 2025 to 2026 is ambitious and assumes both successful integration of acquisitions and timely conversion of contracts.
• Programme risk: defence programmes can be delayed, scaled back or cancelled; even high-quality technology does not guarantee adoption.
• Dilution / overhang: ATM usage, warrants and registered resales can weigh on share price, especially during rallies.
• Competition: both large primes and nimble start-ups are trying to capture the same embodied AI and UAV opportunity set.
• Regulation and export controls: ITAR/EAR and other regimes may limit some international opportunities or create additional compliance costs.

After the Needham Growth Conference and the January 14 slide deck, Palladyne AI presents itself as a vertically-integrated embodied-AI platform with two engines (defence and industrial) and enough cash to attempt a real scale-up. The numbers are still small in absolute terms, but the guidance implies a material acceleration from 2025 and a relatively disciplined cash-burn profile – assuming no major negative surprises on programmes or capital markets.

For a catalyst-driven trader, the story is less about a single binary event and more about a sequence of validation points: named contracts, demonstrable ARR from IQ, visible procurement of Banshee/SwarmStrike and evidence that the manufacturing acquisitions are actually accretive. The company’s own materials explicitly encourage investors and analysts to follow updates via SEC filings, the investor-relations website and its press/news channels, which are the primary sources used for the numbers in this report.

This article does not express any buy/sell view and is not investment advice; it is a structured summary of the latest publicly available information, built to support your own independent due diligence and any future run-up-style strategy notes you may decide to publish on Merlintrader.