Battery life prevents using this for long hauls, but the idea of using these small pods that can easily split up on small industrial spurs doesn't make sense either - most of those non-main-line routes have manually thrown switches. So beyond the much higher unit economics of requiring 2 complex autonomous vehicles (and associated maintenance) per container instead of a basic hunk of steel with wheels, you still need to have a person travel with it to throw switches, or else upgrade every switch on every tiny, poorly maintained spur to be electrically operated.
There's also some strange claims by this company in the article, like the ability for these to carry double-stacks. If you look closely at current intermodal rolling stock, you'll notice that the containers are between bogies rather than above them. A double-stack container would sit too high to fit through standard tunnels and bridges if the containers were on top of the wheels. There's no possible way to accomplish this concept (a pod per end of a container) if you need the container to sit lower than the top of the wheels...
Maybe there's some potential with this idea, but I'm really not buying it based on what they're saying and showing so far.