Ondas is essentially trying to carve out a place for itself where modern defense, autonomy, and physical AI are actually deployed in the field. $ONDS stands out, particularly with its platforms actively used in rail modernization and defense/security operations. But what we're really looking at are the Optimus drone system, the Iron Drone Raider anti-drone hunter, Roboteam's ground robots, Sentrycs' anti-drone solutions, and Ondas Networks' 802.16t wireless network for rail systems… These are being used in the field, receiving feedback from real operations and being continuously improved. Once integrated into the system in defense and critical infrastructure, the learning curve is significantly different and creates a strong stickiness.
The physical AI aspect is what I like most. These aren't just hardware; they're autonomous motors making decisions at the edge. That's why I'm carrying over $AVAV, and its earnings report will be released after the close tomorrow, Monday. Anyway, let's get back to Ondas. Optimus, for example, performs sensor + navigation + operation internally, flies fully autonomously, and collects data. Iron Drone also detects, decides, and neutralizes while in flight. With the acquisition of World View, they have now reached the stratosphere, becoming a layered system. This means that detection, decision-making, and action are all in one package, based on real-world data and processes.
Recent threats and infrastructure needs have clearly demonstrated the critical importance of autonomous drones, counter-drone systems, and dedicated wireless networks. The US is aware of this and will be making 1.5 billion euros in defense industry investments; Ondas will also benefit from this.
The established infrastructure is also a significant advantage. Long-term upgrades are underway for rail communication on several Class I railroads and large systems like Metra. Global customers and partnerships are also increasing in defense (such as the Sentrycs integration with Lockheed). Once a rail operator or security unit becomes accustomed to the system, switching costs become very high, making it difficult to change it. This means long upgrade cycles, recurring revenue, and predictable demand. Management provided guidance of a $457 million backlog and at least $390 million in revenue for 2026, which actually shows us that the visibility side is quite high.
The stock has recently retreated quite a bit from its peaks, which is precisely why I think it's slowly entering the accumulation zone. At least for long-term investors, it makes me think you're not in a bad place to move with DCA. It has everything I'm looking for in my small-cap portfolio: defense and infrastructure spending is growing structurally, not cyclically. Their mission is good, they sell critical products, they have geopolitical momentum, and most importantly, they're already making money thanks to physical AI, taking orders without waiting for regulation or proof.
The stock is expensive according to traditional metrics, so I'm not rushing, but it shouldn't be valued like a classic defense company. It behaves more like a defense-tech autonomy platform; autonomous systems + software layers + AI + data cycles add value over time. If integration continues after acquisitions and margins improve in 2027-28, today's price will be much more reasonable in the future. Of course, if this doesn't happen, it could fall from $7 to $3-5, just as it went from $1 to $15. Knowing this, I'm patiently and slowly building up my position because this company is currently showing us what physical AI really means in the real world. I will increase its weighting in my small-cap portfolio, and if it gets its share in the $1.5 billion defense budget, it will reach the level I want and perform the action I desire.
