Making a Dent
It is rare for new entrants in an established industry to do something completely novel and different from all the incumbents. In a way, that may be the only way to succeed when the established competitors have all the experience, all the experts and all the expertise, all the infrastructure, all the credentials and brand recognition – how are you supposed to survive?
If you’re offering essentially the same product or service, you have to be much cheaper or much better or both, so much so that customers can justify taking a risk with you. Because why would they do that if you are neither?
Being Different
Or you offer something novel and different, which is more likely in some industries than others. In warehouse automation, automatic mixed case palletizing (early 2000s) falls into this category. Shuttle systems don’t, because they were just a variant of what we had already done before. AutoStore’s cube storage solution was also essentially a variation of what we had done before, but it was so much better than other systems in many ways (space utilization, reliability, maintenance costs, planning time, implementation time, business model) while being cheaper (at least initially) that it could be considered revolutionary. Many of the other more recent goods-to-person systems (Exotec, Attabotics, eScala…) would have qualified for this distinction, if AutoStore did not exist.
But it does, so they don’t. Basically, they are all merely variants of goods-to-person systems that have different strengths and weaknesses and levels of maturity, including some aspects that AutoStore pioneered, such as low implementation time and low maintenance needs. The only company I can think of that is doing something exceptional is Instock (listen to my podcast with Yegor, CEO and co-founder of Instock).
With its free-moving, wall-climbing vehicles that drive under the ceiling of intermediate rack levels to stack and unstack totes (you have to see it to believe it, or even comprehend it in the first place), the company is clearly taking a very different approach to anyone else in the goods-to-person solutions domain. For most customers, the technical sophistication and elegance that the storage totes experience on their way to the goods-to-person station is secondary, however. The goods-to-person domain is very crowded, and I maintain that Instock is as differentiated from the rest of the crowd as it gets. I can see that they do things differently, anyone can see that, but my brain isn’t able to tell me if they do things (much) better.
The other day I was lucky enough to have a conversation with Gili Ovadia from BionicHIVE. The more I thought about the system while preparing for the podcast, the more confused I was by what I actually saw. On the surface, I saw an AGV (called “SqUID”) that can also climb up a shelf. In this respect, Exotec does the same thing with the Skypod. Even if we take into account two important differences, namely the fact that BionicHIVE’s SqUID can store and retrieve cartons of different sizes (instead of standardized plastic totes) and can do so in mostly conventional racks (which have been upgraded with “driving rails”, for lack of a better term), the important differences are still not in plain sight.
What is This?
During our conversation, I asked Gili what SqUID was actually doing. He seemed a bit puzzled by my question, and in hindsight I think I understand why, because it’s a bit of a silly question. When I first saw videos of BionicHIVE’s SqUID, it was the second year of Covid, in the midst of government overreach and the flurry of new goods-to-person systems. And I probably registered BionicHIVE somewhere in the vicinity of goods-to-person systems. But that’s not their territory. And this makes all the difference!
SqUID is not another goods-to-person system. It is a multi-purpose tool. Most technical systems in warehouse logistics have exactly one clearly defined purpose, some (few) have two: e.g., most AGVs do transportation, some (a few) can even do storage and retrieval (mostly of pallets, like Linde’s R-Matic). AutoStore, Exotec, Attabotics, Instock, Movu eScala, … do storage and retrieval of totes for goods-to-person picking. SqUID can be used as a storage and retrieval machine for goods-to-person picking (but it wouldn’t make sense as primary application). It can also be used for put-away of cartons right from the goods-in area. It can be used for inventory counting (as a better alternative to warehouse drones). It can perform transportation tasks and can use a put wall in a goods-out area as a destination. It can support consolidation tasks in the same way. It can replenish flow racks at assembly lines or zone picking systems. And it could do single-item picking on the rack directly at the storage location if it is equipped with an additional picking arm.
Now, most of the things that I said SqUID “can” do, it actually can’t do today. But there is not reason why it should not be able to do those things in the future. There are no conceptual limitations, and that’s what makes the system so interesting. The potential is huge, and if the guys at BionicHIVE manage to stay afloat and maintain their edge, a system with this versatility could truly be a revolution in warehouse automation.
Challenges
It is not going to be easy, though. Some very practical problems need to be solved, and it is not self-evident whether they can be solved without changing some of the core features of SqUID. Mixed operations are one of these features, and safety is one of these problems. In the podcast, we did talk about aisle safety (min 34:48). I learned that robots cannot drop, whether they lose power or for other reasons. So, provided the rails are properly attached to the rack (which I suspect will have to become part of regular safety inspection), there is very little risk a robot would fall onto someone walking the warehouse aisles. However, there is still the risk of boxes, or contents of boxes, falling. The risk is not necessarily high, but it is higher than zero. I would imagine that in some countries, this could make getting approval for the system difficult. Gili mentioned they are working with authorities in California (OHSA) to figure out regulations. California is not exactly my home turf, so I’m not qualified to comment on it, but based on hearsay I think the level of regulation there is comparatively high. If the system gets approval in California, chances are it will get approval in other places, too. If it doesn’t, the party is over before it really began. This is not a technical challenge (of which I am sure there are many), but a conceptual one, which is much harder to solve.
There are other challenges. As a start-up with ambitions to grow, BionicHIVE will need capital. Besides the chicken-and-egg problem we touched upon in our conversation (investors want to see customer projects before they provide funding, but you need funding to secure customer projects…), we know that profit margins in the warehouse automation projects are not particularly high. To generate a return attractive to venture capitalists, you need a lot of projects, and especially when your projects are rather lightweight, as is likely the case for BionicHIVE. Lightweight projects (= low total CAPEX) are advantageous for many reasons, but their returns in absolute terms are also relatively modest. Again, this is not a technical challenge, nor is it unique to BionicHIVE; it is a challenge that all start-ups in warehouse automation have to deal with. To my knowledge, the only company that is generating really high returns is AutoStore. Part of their success is due to their business model: they delegate the risky project business to integrators while focusing on producing a low number of highly standardized components, which is mostly risk free. So, it is no wonder that investors put in large amounts of money and took the company public. However, the development of their stock price does not indicate market enthusiasm. To be fair, though, since going public they have experienced the burst of the tech hype bubble and then got schooled by Ocado’s lawyers, and both impacted their stock price.
Conclusion
In summary, BionicHIVE’s SqUID could be a potential game-changer in the warehouse automation industry, not by being another goods-to-person system, but by being a versatile, multi-purpose tool. This unique capability could revolutionize how warehouse operations are conducted, provided the company can overcome some challenges. These challenges are not merely technical but conceptual, most notably involving safety concerns and the broader issue of gaining regulatory approval in highly regulated markets.
Furthermore, BionicHIVE, like many start-ups, faces the inherent difficulties of securing capital and demonstrating profitability in an industry with typically thin margins and high project risk.
I do think, however, that the the SqUID system offers a compelling vision. At this point, it is the only system that shows us how robots and humans could work together in warehouses beyond the narrow confines of picking-only or transportation-only tasks.
Podcast
Please find below the recent podcast episode that I’m referring to in the text.
