Tristan d’Orgeval, Chief Strategy Officer and Co-founder
“As Nomagic closes its Series B fundraising, our plans to transform warehouses with AI-powered robotics and automate repetitive, manual tasks are stronger than ever. This revolution we are driving, together with our customers follows another series of profound changes in intralogistics over the last 30 years.”
First, looking back
Over recent years warehouses have seen the introduction and expansion of various forms of goods-to-person systems:
Shuttles and miniloads, delivering performance for large customers
Cube storage systems such as AutoStore and Ocado, densifying storage
AGV based solutions such as Exotec or Hai Robotics bringing scaling flexibility
AMR based solutions such as Locus, GreyOrange or Geek+, allowing easy retrofit from fully manual operations.
Warehouses have also seen progressive automation of sorting and packing with the introduction of various sorters (pocket, line, tilt-tray, etc…) and packing machines, from height reduction systems (e.g. B+ machines) to fully fitted packing (e.g. CMC or Packsize). They have seen the continuous expansion of conveyors, AMRs and all required controlling systems for the “simple” transportation of bins and cartons.
This leaves intralogistics in a relatively diverse state, where even today a number of operations remain fully manual. However, most warehouses have introduced some degree of automation, from a single central goods-to-person or sorter, to multiple systems interconnected with only limited touch points by human operators in between.
Where is the hype?
With this diverse state, where is the hype? Probably 3 things come to mind:
Humanoid robots – It all started with Boston Dynamics and their robot Atlas. Now discontinued… But the promises brought by recent AI advancement combined with the science-fiction feel of some humanoid robot videos definitely makes it a hype of the moment. In warehouses? Not so much yet, but companies like Agility Robotics want to make it a reality.
AutoStore competition – As AutoStore managed to establish itself as the de-facto choice for reliable storage systems, appetite grew to compete and is now filled with promising solutions. The cake is large and it seems no one is ready to accept that it would be a winner-take-all market, so hype is probably here to stay for new storage systems like Exotec, Hai, Movu and Brightpick,…
AI-powered stationary robots – With a number of warehouses having eliminated the need for operators to walk around using goods-to-person or sorters, what remains is manipulation tasks, typically pick and place, hard to automate because of the variety of products and packages to handle but essential to address to get to fully automated (and therefore more predictable) workflows.
Finally, looking ahead
Now this gets us back to our title. Warehouse automation in 2030… What will that look like? Some tentative predictions:
Humanoid robots will remain cool (and borderline useless) – What will prevent humanoids from becoming THE key to warehouse robotic fulfillment is this truth in automation: To automate a process, you should not copy how humans do it today. Why? Because it’s most likely not the most cost and energy efficient. This ties back to Taylorism. If you can divide tasks, you will be able to make each one much more efficient and in the end design a highly cost & energy-effective machine for it. So, wheels and by extension, AGVs, Conveyors are far better than legs and arms to transport heavy goods. And once they are transported, why would you still need legs to just pick and place those goods?
Lights out warehouse will become a reality (and Amazon will not be first) – The underlying AI technology used for humanoid robots and its progress will however be key to expand AI-powered stationary robots, enabling them operate at any point of the chain where a manual touch point still exists. Automation of every touch point will be easier to achieve when the variety of items is relatively low and in warehouses where processes are less diverse. With their millions of SKUs in each warehouse, this makes Amazon a less likely front runner on this one. If I had to bet I’d say highly innovative companies in Fashion or Pharma stand a good chance there. First, the highly competitive and regulatory pressure in these industries, combined with rising labor costs will make lights out warehouses a game changer for them. Second, their relatively standardised assortment (soft garments for fashion, small boxes and bottles for pharma) may lend itself well to full automation.
AI-powered manipulation robots will become the differentiating factor – As leaders in every industry race towards lights out warehouse, the missing piece in automation will be this touch point or manipulation required in between already well automated systems. Companies adopting this technology early, adapting their workflows and leveraging the progress on AI driven Robotics will outpace others in their ability to fulfill orders at scale, without disruption and at a progressively lower cost while labor costs will continue to rise.
New systems to oversee and orchestrate automation will emerge – With the realisation of lights out warehouses, the number of systems to coordinate end-to-end autonomous fulfillment will grow massively, including vision and video based quality control systems. Such systems will have a hard time integrating smoothly into rigid Warehouse Management and Warehouse Control Systems, so they will have to develop their own separate networks and architecture to give the Operations Directors the visibility and comfort they need to run those fully automated systems.
These predictions are some of the elements driving our vision at Nomagic and where we’ll put our focus for the next 5 years. We’ll be scaling AI-powered manipulation robots throughout Europe and the US to support this vision and help our customer achieve higher warehouse performance.
I hope to engage with the great minds and operators in this industry over the next 5 years to help make this vision a reality!
