Share this
by jpallmerine on 19 January, 2023
Key Takeaways
- Many warehouses still run core systems on local servers, not the cloud.
- AutoScheduler.AI maps each site as actors, actions, constraints, and flows.
- Trust builds slowly, so early customer work shapes the whole rollout.
- Automation only works if people can explain the process to it first.
Automation is reshaping almost every industry, but algorithms still need people to point them in the right direction. That's as true for dock scheduling software as it is for full warehouse orchestration.
Keith Moore, then Chief Product Officer and now CEO of AutoScheduler.AI, works with some of the largest companies in the world to balance automation with existing supply chains. In this first episode of What's Up Dock?, he shares his view on where warehouse automation is today and where it's headed.
Join us as we discuss:
- Designing automation to fit large-scale operations.
- Technology development and maintenance.
- The human side of automation.
How Do You Design Automation to Fit a Large-Scale Warehouse?
Large-scale automation starts with knowing exactly what data you have, what you still need, and what a good project looks like. Most companies have more data than they know what to do with. The challenge is turning it into decisions. Keith starts every project with four questions:
- What data do we have?
- What data do we need?
- What does a good project look like?
- How can we use artificial intelligence?
Once data is collected, it won't tell the business what to do on its own. Think of it as a raw material that has to go through several processes before it's useful. Keith breaks data strategy into three tiers:
- Visibility. Can you use the data to find connections?
- Predictability. Can the data help predict what happens next?
- Prescription. Can the business change direction based on those predictions?
Picture a supermarket that discovers customers tend to buy bananas and grapes together. It can predict the pattern will continue next year, but the real value comes from deciding what to do about it. Knowing where your costs actually sit is part of that first step, which Uponor's Tom Rowbotham digs into in episode three, on analyzing the cost of your supply chain.
Why Do So Many Warehouses Resist Automation?
Many warehouses resist automation because replacing a core management system feels risky and disruptive, and many haven't even moved those systems to the cloud yet. Keith's biggest surprise during the COVID-19 pandemic was how many warehouses still ran their management system on a server sitting under someone's desk.
"We break every site into a series of actors, actions, constraints and flows. We drop all that in a massive configurable algorithm, and then the optimization can run." (Keith Moore)
To get there, Keith's team pulls in a substantial amount of data to map a site's full layout into AutoScheduler.AI. That supercharges the operation, but it requires a lot of trust. He compares it to ripping out all the plumbing in your house for a new system: the company has to understand why the change is essential, not just a minor improvement.
Why some organizations keep clinging to outdated supply chain methods is the focus of episode two, on how technology accelerates logistics, with INXEPTION's Roberta Tamburrino.
"You have to be in some of the first engagements with customers. So we can get a real feel for the operation and draw out all the different inventory flows through rebuilding." (Keith Moore)
What Role Do People Still Play in an Automated Warehouse?
People still define the process automation is meant to run, because a system can only be as smart as the information it receives. Once AutoScheduler.AI is in place, the work isn't over. Keith stresses that organizations need to understand their own needs precisely to get real value from AI.
"If you can't explain it as a person, there's no way you can expect any automated machine learning system to do it well." (Keith Moore)
A process that looks simple can be far more complex step by step. Take unloading crates from a truck: who moves them, where do they go, and when does each step happen? By the end, there might be 100 steps just to unload the product, and every one needs to be understood before it can be automated.
Tools like a yard management system help by making each trailer move visible, but someone still has to define how the work should flow. Products that are heavy, perishable, or packaged several ways add even more steps, as our episode on the logistics of beer shows.
Keep connected with us by subscribing on Apple Podcasts, Spotify, or anywhere you get podcasts. Find all the episodes here.
Whether it's a warehouse-wide algorithm or a single dock door, automation earns its keep only when someone has already mapped out how the work should flow.
Frequently Asked Questions
Who Is Keith Moore?
Keith Moore is CEO of AutoScheduler.AI, a warehouse orchestration platform, and was its Chief Product Officer when he joined What's Up Dock? He works with large companies to balance automation with their existing operations and explains why trust between software and its users takes time to build.
Why Do So Many Warehouses Still Use Manual Systems?
Many warehouses stick with manual systems because replacing a core management system feels as disruptive as ripping out the plumbing in a house. Keith says the real barrier isn't the technology itself, but convincing an organization the change is worth the disruption.
What Data Does a Warehouse Need Before Automating?
A warehouse needs to know what data it already has, what data it still needs, and what a good automation project looks like. Keith breaks data strategy into three tiers: visibility into connections, predictability of future outcomes, and the ability to change direction based on that data.