Former Tesla lead unveils warehouse robot that can lift 3000 lbs

The rise of foundation models has reignited interest in robotics. From Boston Dynamics to Figure, companies are exploring how continuously learning humanoids can transform enterprise and consumer ecosystems. But does it always have to be a complex, two-legged robotics system roaming around? Well, Chris Walti, the former head of Tesla’s Optimus humanoid project, thinks otherwise, especially in the case of warehouse operations. Today, he announced Mytra, a startup co-founded with former Rivian lead Ahmad Baitalmal, with a mission to use simple, AI-driven, 3D robotics systems to handle one of the biggest tasks within a warehouse: material flow.

“We’re taking a radically different approach by reducing the number of parts and moving the focus from hardware to software. We are the first and only solution that can universally automate many of the most labor-intensive, costly, and complex aspects of material flow, which are the ‘red blood cells’ of any industrial operation,” Walti said in a statement.

The company has been operating under the radar since 2022 but is finally emerging from stealth with $78 million funding up to series B. The investment has come from some major industry heavyweights and venture capital firms, including Greenoaks, Eclipse, Abstract Ventures, Promus Ventures, LifeX, 515 Ventures and Y Combinator CEO Garry Tan.

What exactly does the robotics startup provide?

Material flow, or the movement and storage of materials, products and goods, is the lion’s share of work in a warehouse. However, even with all sorts of advances in AI and robotics today, many enterprises handle it with humans driving forklifts or handling conveyors. This is not only time-consuming but also unsustainable, given the increasing pressure on manufacturing and the supply chain as well as the dwindling labor supply.

Walti saw this problem firsthand during his seven-year-long stint at Tesla. He tells VentureBeat that the situation of the material flow system back then was so bad that it “almost sunk the company.”

“I created and led an internal team to ensure Tesla never was in this position again. As the scope of this effort increased, including leading the Tesla Bot, it was clear that more than ‘general purpose’ robotics, the world needs better, simpler, form factors that can harness the power of modern software to solve highly pervasive problems in factories and warehouses,” he said. This ultimately led to the inception of Mytra in 2022.

At the core, the startup provides enterprises with a modularly deployed cell-based matrix structure through which an AI-powered bot navigates. It can carry up to 3000 pounds from any cell to any adjacent cell in any direction, given the needs of the warehouse, delivering a significant upgrade over manual ways of lifting, moving and storing materials. 

“The structure can scale in any direction and form to any shape. Technically, it’s a tesselating lattice structure. It has no moving parts or powered devices and serves to ‘structure’ a 3D space for the bot to move around from any cell to any adjacent cell in any axis (X, Y, and Z). Together, these two systems enable the physics-limited maximum kinematic flexibility,” Walti explained.

Meanwhile, the end-to-end AI software enhances the bot’s efficiency, optimizes routes and manages inventory. It uses manufacturing facilities’ predicted plan (based on past data) of what needs to be done and places inventory that is high velocity or needed soon in strategically higher throughput zones in the structure. Meanwhile, if the goods are not needed soon, it stores them in lower throughput zones.

“Every cell can be software-addressed as a storage or throughput location. This way the Mytra system creates nearly infinite configurability where goods are stored and, conversely, the paths through the system are defined,” the CEO added. Essentially, depending on needs, teams can modify the robotics’ system for the required scale, density, throughput and application at any point in time.

To meet the computational demand of this software-based approach, Mytra has deployed a combination of algorithms trained with reinforcement learning.

“Algorithms focus on small subsets of the problem which are critical and high-value add. The reinforcement learning addresses this computational complexity in a very elegant and efficient way by sorting through the limitless possibilities to create an efficient plan for the operations at any point in time,” Walti noted.

Significant traction from Fortune 100

While the company has just emerged from stealth, the founder claims they are already seeing significant traction from Fortune 100 companies like grocery giant Albertsons (which has also invested in the startup).

“We have pilot programs in progress and interest from existing customers who want to scale their work with us. More recently, we’re seeing growing demand from new customers, with several new deployments slated for 2025. The strong traction we’ve seen so far from customers demonstrates we are solving a critical and growing business challenge,” Walti added.

When asked how much deploying the modular system would cost, the CEO emphasized that the “upfront investment in the structure should be comparable to traditional pallet racking at scale.” The other elements of the system – the bots and AI software – are offered on a service-based model.

“With all of our projects so far we have seen significantly lower costs of using Mytra bots + software vs. traditional material handling equipment (forklifts, pallet jacks) + human labor for the same activity,” he said. 

Notably, companies like AutoStore, Locus and Zebra are also working to automate material movement tasks with the help of robots, the CEO claims that all those solutions require deep technical expertise and a massive investment of time and money. On top of that, they lack the flexibility to adjust to evolving business needs.

With the funding, the CEO plans to scale the technology and deliver the system to interested customers. It will also grow its team across departments to further improve the system and deploy the next generation of Mytra bots for blue chip companies looking to automate applications that remain highly manual and labor-intensive.