Blog

The Bot Company Shortlist: 6 Robotics and AI Companies Worth Watching

The six robotics and AI companies most worth watching are The Bot Company, Figure AI, Agility Robotics, Skild AI, Physical Intelligence, and Covariant. Demand is not theoretical either: global industrial robot installations hit 541,302 units in 2023, the second-highest annual total on record [source: IFR].

TLDR: These six firms matter because they are pushing robots out of demos and into homes, warehouses, factories, and general labor tasks. A warehouse handling 10,000 picks per day could save serious money if an AI robot cuts mispicks by even 1%, since that is 100 fewer wrong orders daily. The best companies in this group are not just building arms and legs. They are building robot brains that can learn, adapt, and work with less hand holding.

The shortlist: 6 robotics and AI companies worth watching

Robotics has had false starts before. The demos looked great. Then the robot met a messy kitchen, a shiny warehouse floor, or a half-open box. Trouble started.

The newer wave feels different because AI models are getting better at vision, instruction following, and task planning. Still, nobody should pretend this is easy. It drives teams crazy when a robot handles 95% of a task, then wastes 40 seconds failing on the simplest edge case.

1. The Bot Company

The Bot Company is one of the most watched names because it is targeting useful household robots. That is a brutal market. Homes are messy, personal, and full of odd objects. A robot cannot assume every room looks like a lab.

The company stands out because its goal is broad home assistance, not one narrow gadget. If it succeeds, it could move robotics from factories into daily life. That would be a huge shift. The hard part is trust. A home robot must be safe, quiet, useful, and affordable. Missing one of those marks can kill the product.

2. Figure AI

Figure AI is building humanoid robots for physical jobs. Its robots are designed to work in spaces built for people, which makes sense. Stairs, shelves, doors, tools, and factory stations already fit human bodies.

Figure is interesting because it is aiming at labor gaps in manufacturing and logistics first. That is smarter than starting with a cute consumer robot. Businesses can measure return on cost faster. If a humanoid robot can move parts, carry bins, or support assembly work for long shifts, buyers will pay attention.

The catch is that humanoids are difficult. Balance, grip, battery life, and safety all have to work at once. A robot that looks impressive in a video may still be too slow for a real factory line.

3. Agility Robotics

Agility Robotics is best known for Digit, a biped robot built for warehouse and logistics work. Digit has a practical shape. It can walk, carry totes, and operate in existing facilities without forcing every company to rebuild its floor plan.

That matters. Warehouses hate disruption. A robot that needs months of layout changes is a hard sell. Agility’s focus on material movement gives it a clear problem to solve. That is better than vague talk about “general robots.”

Its challenge is speed and uptime. Warehouse managers care about boring numbers: tasks per hour, downtime, repairs, and training time. If those numbers beat human-only workflows, Agility could become a major name.

4. Skild AI

Skild AI is building AI models for robots rather than focusing only on one machine. That position could be powerful. Many robots fail because the hardware is fine, but the brain is too rigid.

Skild’s pitch is a general-purpose robotics model that can work across different bodies and tasks. If the model can transfer skills from one robot type to another, it may cut training time and development cost.

This is a big “if.” Robot learning is harder than text or image AI because the real world pushes back. Objects slip. Sensors miss things. Floors change. Still, the company’s model-first approach is worth watching.

5. Physical Intelligence

Physical Intelligence is another company focused on AI for real-world robot control. Its work centers on getting machines to understand physical tasks through data and learning.

This type of company may become the hidden engine behind robotics. Not every winner needs to sell a famous humanoid. Some may provide the intelligence that lets many robots become more capable.

The annoying part is that progress can look slow from the outside. A better grasp, smoother motion, or lower failure rate may not feel flashy. Yet those gains matter. A robot that fails once every 20 actions is a demo. A robot that fails once every 2,000 actions starts to look like a product.

6. Covariant

Covariant focuses on AI-powered robotic picking and warehouse automation. Its systems help robots identify, pick, and sort a wide range of items.

This is a practical market. E-commerce warehouses deal with unusual shapes, packaging, reflections, crushed boxes, and constant SKU changes. Traditional automation struggles there.

Covariant’s strength is that it attacks a job with clear value. Faster picking means better throughput. Fewer errors mean fewer returns. For retailers and logistics firms, those gains can show up directly in margins.

What makes these companies stand out?

  • Real use cases: They target labor, logistics, home help, and physical automation.
  • Better AI: They focus on perception, reasoning, and control, not just metal parts.
  • Large markets: Warehouses, factories, and homes all have room for useful robots.
  • Hard problems: Each company is working on tasks that older automation could not handle well.

FAQ

  • Which company is best positioned right now?
    Agility Robotics and Covariant look closest to clear business use, especially in warehouses.
  • Which company is the boldest bet?
    The Bot Company may be the boldest because home robots are so hard to build and sell.
  • Are humanoid robots ready for mass use?
    Not yet. They are improving, but cost, speed, safety, and reliability still need work.
  • Why are robotics companies getting so much attention?
    AI has improved robot vision, planning, and language understanding. That makes physical automation more realistic.
  • What should investors or buyers watch first?
    They should watch uptime, cost per task, deployment time, and failure rates. Flashy demos matter less than repeatable work.
To top