Warehouses are a critical part of nearly every industry in the world. Technologically-speaking, however, they lag far behind most of the sectors they serve. Where other industries are powered by software and automation, the crucial waypoints of the supply chain still rely on manual labor processes and complex, heavy machinery. Today, 95% of warehouses still have no automation.

While this gap may have been workable even just a few years ago, it’s a different story today. Warehouse-reliant businesses are changing how they order and store inventory to accommodate shifting consumer demands, material shortages, and geopolitical uncertainty. Keeping up with these new strategies requires a level of visibility and efficiency that low-tech warehouses just don’t have. 

In the meantime, humans are stuck doing repetitive, labor-intensive jobs that not only lead to injuries and burnout, but also take up entirely too much time, cannot easily adapt to the rapid changes in inventory or shipping schedules, and are an impediment to scaling. All of this is compounded by historic labor shortages that are projected to get even worse — by 2030, experts estimate more than 2 million manufacturing jobs in the U.S. will go unfilled. Yet for most warehouses, a low supply of humans is better than any of the tech solutions offered today, which mostly consist of point solutions that don’t integrate into increasingly complex warehouse workflows.

Instead of looking at the various small, detailed jobs within a warehouse, Tesla alums Chris Walti and Ahmad Baitalmal looked at the basic functions that make up the largest share of the workload: moving materials. Some 80% of the work in warehouses (and 40% in manufacturing environments) is devoted to finding, retrieving, moving, and storing goods. As former engineers at the EV pioneer, Walti and Baitalmal were deeply familiar with the pain of managing material flow in industrial environments. They also knew that even if more human beings (or humanoid robots) were available, they might only complicate things further. So they looked at it from a first-principles angle: instead of more humans to move the mountain of materials from one place to another, what if the mountain could just move to the humans? 

And so, Mytra was born. Co-Founded by Walti and Baitalmal in 2022 in partnership with Eclipse Venture Equity (VE), Mytra’s AI-powered, robotic system effectively turns every pallet of material in a warehouse into its own intelligent, efficient, and extremely strong elevator of sorts. The AI-powered system consists of a robotic pallet, a customizable tray, and software. Using advanced computer vision, sensors, and IMU feedback, each bot functions independently — and can move in six cardinal directions while lifting up to 3,000 pounds.

In July 2024, Mytra came out of stealth with $78 million, and the company’s impressive traction illustrates the scope of the problem – and the massive opportunity –  they are targeting.The company scaled up the first generation of their AI-powered, robotic system within the first two years of their existence, built their first testing site, and has secured partnerships with several Fortune 100 companies —  including the second-largest grocery chain in North America, Albertsons.

“We’ve Been Our Customer”

Instead of attempting to build a product tailor-made for the unique complexities of specific manufacturers and warehouses, Mytra looked at what was the same in every environment: the pallet. 

“Picture a warehouse. It doesn’t matter which type of warehouse and at which point in time in the last 100 years – it will look the same,” says Baitamal. “That’s why you need a solution that gets to the core of what powers a warehouse: the pallet. The pallet is the red blood cell of the system. It moves everything.”

Baitalmal, who previously led factory software at Tesla and Rivian, said finding product-market-fit came from his and the rest of the founding team’s direct experience. They saw not only the impact of low tech penetration in manufacturing environments every day, but how the root challenges couldn’t be addressed with software-focused point solutions. 

“All of us have worked shifts in warehouses and factories. We have the same incentives as our customers because we’ve been our customers,” says Baitalmal. “To really make a difference in warehouses, you have to stop thinking in 2D. You only have to walk around a 100,000-square foot warehouse once to see that these are 3D problems requiring 3D solutions.” 

Where others may have envisioned a 3D solution as a fleet of humanoid robots, Walti knew from his previous experience as the lead for Tesla’s Optimus Bot that this technology was still years away. He and Baitalmal knew that attainable innovation had to involve a systems-level approach. The result is a modular, customizable system that can handle up to 60% of the material flow tasks — a game-changer for pretty much any environment where the core function is moving materials.

“The main question to answer, over and over, was whether what we were trying to build broke the laws of physics,” said Baitalmal. “As long as it’s not doing so, we just have to keep perfecting what works, and keep making it simpler. After that, the only thing that could ever hold us back is self-doubt.”

Achieving that in a format that can be widely applicable to a range of industrial environments means keeping those North Stars in sight: make a product with as few components as possible, operate it with a simple technical interface that anyone can use, and ensure customers can easily deploy the system without causing disruption to their existing operations. The vision was simple. Bringing it all to fruition was not. 

“The most important thing is integration — not just of the different parts involved here, but with the people building them,” says systems engineering lead John Paton. “You have to be really comfortable with the fact that you are going to have to experience a lot of failure in order to learn what works.”

Seven Disciplines, One Vision

The multibot, material flow system may not seem as complicated as a humanoid robot, but Mytra is far from a simple product. 

“We have seven different engineering disciplines under one roof, and we’re all working together to make something completely new,” says Walti. “There’s no way to do that unless we are all in alignment on the vision for what this product is fundamentally supposed to do.”

The background of the lean — but mighty team — at Mytra spans robotics, electrical engineering, structural engineering, AI, systems engineering, manufacturing, go-to-market, and more. Walti says that breadth of expertise across teams (some of which were just one or two people in the beginning) could only be fully realized towards a common goal with near-constant communication. That meant daily standup meetings across the entire company, frequent one-on-one discussions, extensive documentation, and a lot of hustle across the team. 

In their first year, Mytra went from blueprints to prototype, testing a small scale, 3-cell version of the assembly in their own South San Francisco facility. In the second year, they began testing in the large, 36-cell assembly  in Tracy in partnership with Albertsons.

Now, the company is currently working on the second generation of the system, having achieved the critical physical milestone that makes the whole thing possible: the ability to move in all six cardinal directions while carrying thousands of pounds of materials.

“It never seemed impossible, but getting the robot to move in a Z [with that heavy of a load] hadn’t been done before,” says Rachel Leu, a senior electrical engineer who’s been with Mytra since the beginning. “People were nervous, and it was really hard to get right, but when it happened, it was a defining moment for all of us.”

Like many at Mytra, Leu was drawn to the company for the opportunity to work with a diverse group of people hailing from unique backgrounds, and to build something entirely new. Figuring out how to do that in the quickest way possible meant utilizing a mix of basic off-the-shelf components, like power control panels and circuit boards, and combining those with novel creations. 

Those early iterations didn’t always look or function ideally, says Leu, but they were crucial for the learning process.

“Sometimes you have to put things together that are a little bit janky just to the point where you can learn, teach the rest of the team, and show customers how it works,” says Leu. “Now we’re really pushing for the next stage.”

The bright letters that spell out “DEJANK FACTORY” above Leu’s workstation signify work on Gen2 is officially in full swing. Now, Leu and the rest of the Mytra team are moving away from off-the-shelf components and building parts tailored to their specific needs in-house, such as power control panels that can mitigate harnessing problems of generation 1, better sensing inputs, and stronger power distributor units. 

Software engineer Sasha Rudolf described a similar learning process to get between Gen1 and Gen2 when it came to building the control functions of the bot system. Since the success of the Mytra system rests on properly bringing together several individual disciplines in one product, it was crucial to learn how they all need to work together before committing to custom builds, Rudolf says. Any problems discovered in bespoke systems would just require rebuilding everything all over again, which isn’t feasible when working with extremely tight deadlines. 

“There were moments where we were days away from a critical deadline before we finally got it right, such as figuring out how to implement sensors and a kill switch to prevent tilting,” says Rudolf. “But we set ourselves up for success because of the way we learned – even if it didn’t make sense at the beginning.”

Beyond Warehouses

In just two years, Mytra has essentially reinvented its product, landed critical partnerships, and doubled its team. Still, they are only at the beginning of their journey.

“Now, everywhere I go, I see pallets everywhere — hospitals, airports, farms — everywhere,” says Baitalmal, describing the future applications of the Mytra system. “As long as it is an environment where we are moving material, we can improve things.” 

Today’s pressures on the supply chain come from a convergence of factors – rising shipping costs, a shrinking labor pool, and higher expectations from customers whose demands are increasingly difficult to forecast. These trends aren’t fleeting, and are only projected to get more pronounced, especially in the U.S. – with an incoming administration that has indicated a future with tariffs and regulations that favor domestic manufacturing and materials sourcing, warehouses will need all the help they can get to keep up. Automation is critical in order to build resilience and efficiency into the supply chain system, and Mytra aims to be the leading provider of vertically integrated software and hardware for more than just warehouses.

“These are tough workplaces with a lot of turnover,” says Baitalmal. “Mytra doesn’t replace anybody, it makes their life easier by reducing the time they are working in harsh environments and giving them more time to think strategically about how they run their business.”

Follow Eclipse on LinkedIn or sign up for Eclipse’s Newsletter for the latest on the Industrial Evolution.