What is a post office? You might say it is a place to post letters. That works when people post letters. But the post office also delivers letters and sells stamps. It runs fleets; both inbound and outbound. A post office is a bank, community outreach center and a collection point. It’s a hub.
When we apply the thinking of commercial drone vendors (a recent blog post topic) to the Post Office model, we look at the tool first. The people who develop drones want you to take hardware, autonomy stack and fleet management tools from one vendor, them. Often, they’re selling “complete” solutions. The problem is the word: Complete.
The trouble with that concept is that we’ve already had it in the IoT sector. The client hates it. It is called vendor lock-in. If you can think of drones as “interchangeable hardware layers” you’re onto a better proposition, a more solid platform. If you can map the control layer as an independent system layer, you can begin to build in different modalities, suited to the specifics of different sectors.
With two simple changes, the client can swap vehicles, upgrade device autonomy and change vendors. Just like the post office, the flexibility to integrate different functional elements is valuable. Whoever owns the routing, scheduling, optimization and data layers owns the system. This is why the post office is iconic, institutional (…and up until recently, effective).
Hubs
Cargo systems don’t work because of vehicles. They work because of defined handoff points (the church, post office, lending library and village shop; the archetypal community). In our drone analogy, we’re looking at loading bays, charging stations, battery swaps and maintenance nodes. If those are proprietary you’ve got lock-in. If they’re standardized you’ve got an ecosystem. Standardize the node, not the machine.
Energy is the hidden lock-in vector. This is why we see such fragmentation in the e-bike and e-scooter rideshare sector. It is why macro level entities like the EU, or Germany, want to standardize batteries, to avoid sector wide waste, duplication and service issues. We’ve seen the mess that cowboy operators can manage in pursuit of sectoral dominance.
Reversible system design anticipates partial failure
One vendor can fail, one vehicle class can be removed and one node can go offline. But system wide the network can handle it. It is why “everything works together perfectly” is not a structural imperative. We design contingencies for failure. Unmanned cargo vehicles are distributed physical elements of a communication network. No single vector should control the whole. The smart play is in building critical architecture (things that must work) that stops a system from becoming a cage.
