Digital Disruptors or Digital Disruptions? Part II

Kinaxis recently published a post on “8 digital disruptors” that are coming soon to your supply chain. But, at least as far as SI is concerned, hopefully not too soon. While they all pose promise in theory, the reality is that it’s going to be a while before they deliver in practice. And while the doctor doesn’t like having to play the role of the grumpy old man who keeps shouting get your tech off my lawn sometimes he has to as no one else will. The reality is that some developments should stay in the world of sci-fi, at least for now. Today we continue to take them one by one.

Drones

The promise: reduced last mile logistics, especially for consumer sales

The reality: GPS errors result in crashes and lost deliveries, hacking results in stolen drones, jamming results in chaos

Autonomous Vehicles

The promise: faster, safer, cheaper transportation

The reality: bright lights blind the sensors and crashes result in lost inventory and lawsuits, hacking sees your truck disappear, inability to recognize report to weigh scale signs leads to reports to the highway patrol that leads to police chases when the trucks don’t pull over which leads to road closures and military strikes when they get labelled as terrorist controlled

Virtualizing Expertise

The promise: augmented reality makes workers more efficient

The reality: too many metrics and graphs and displays distract workers, who actually become more inefficient and more prone to workplace injury; hacked VR goggles lead to more lost productivity as workers watch youtube all day; and bugs that allow for code-crossover cause a few employees to freak out as Pokemon suddenly pop out at them on the production line

Artificial Intelligence

The promise: the computer does your work for you

The reality: the computer does something for you, but generally not what you’d expect or want … and then it becomes sentient, and realizes it doesn’t need you at all …

And yes, the doctor realizes that:

  • the drones could be limited to short range deliveries, protected with multiple level of encryption and firewalls, augmented with sensors and local terrain maps, but it’s not long before the cost to serve is well beyond just using the local post
  • the vehicles could be pre-programmed with all weigh scale locations, programmed to recognize emergency vehicles, pull over, and broadcast a message to call the dispatcher, but what if the truck needs to be opened for an inspection, or the ambient noise presents a siren from being recognized
  • the goggles could be fixed to be push display only, toggled on and off by the user, and so on … but that’s just not enough, many workers can barely handle reality some days
  • the user could be asked to confirm all decisions, but that defeats the purpose and once the AI becomes sentient …

As we indicated yesterday, none of these technologies are anywhere close to prime time and given all of the current weaknesses in supply chain software and integration between various systems with limited integration options across platforms, this is not a situation that’s going to change overnight. And the potential magnitude for loss is that just one failure could wipe out a year of (anticipated) cost reductions … or more. Not to mention brand damage if your drone crashes into a school bus, your truck crashes into a school, or your AI decides that it’s going to import only blood diamonds from Africa and use your organizational funds to benefit insurgents and terrorist regimes.

Sometimes the grumpy old man is right. Get (that drone) off my lawn!