Fukuoka: hyper-small villages and automated fruit transport

Short update for Fukuoka that new automated driving tests will begin on a 5km stretch of road between the citrus-picking area of Ichinoki, and the Ichikawa town hall representative office at the local “roadside station”. Location of the test: the small village of Miyama (population: 103). Period: from September 2nd through September 9th. I think that qualifies this village for the smallest in Japan to run automated driving tests in our research so far, but as I’ve said in other regions for similar tests, I believe one week is far too short to make solid statistically-relevant conclusions… in any case it’s a start.

Plus there are two additional interesting characteristics of the tests I haven’t seen before. One: literally zero public transportation in the area. So these tests will enable a direct comparison with vehicles driven by local residents, ie something like an A/B test with no mobility mix to reference. Two: this village is one of 13 new locations signed off by the Ministry of Land and Transport, but its size is by far the smallest and a possible model test case for hyper-small rural community automated mobility across Japan.

Going a step further if one considers the growing problem of low-population-density rural sprawl, especially in the vicinity of tier-2 rural prefecture cities, one could think of modelling mobility in this sprawl as a network of hyper-small rural communities tied together by the rural road network. Unfortunately it is highly unlikely that there will be “roadside stations” for each hamlet of 103 souls, but one could define zones of regular passenger/car footfall transition as virtual roadside stations in the system. These could be gas stations, mini-/mega-malls, linear collections of shops hugging the roadside, bus stops, warehouses/depots and so on.

This is especially important if the idea in the future is to employ automated vehicles to transport citrus fruit from the hills of Ichinoki to not only the local roadside station where they may or may not be enough footfall to warrant transportation of goods, but further afield. It is more likely that in many cases the vehicle will go straight to the larger conurbation’s closest depot, supermarket or mall to drop off goods and ensure the trip was economically viable. Route optimization would then become a balance between contracted volume routes (eg. 50kg/week of fruit contracted by a local supermarket), spot pricing (one-off requests, urgent demand, warehouse stock hedging, local restaurants etc), any loss-leader/regulation-required drop offs (eg min 5kg/wk of fruit at local roadside station for PR/tourism purposes) and available stock of produce.

On the passenger side, one could update these virtual roadside stations/stops as passenger demand requires it. Citizens using the system could voice their concerns directly to the vehicle. This could happen via a voice-activated system with a mic and a button built into the vehicle that allow any concerns to be stored, sent over the cellular/wifi network when available, converted from speech to text automatically server-side, and stored in a database where the concerns and the requested stops could be automatically analysed for addition/removal of the network on a semi-automatic basis. This could provide the benefit of a somewhat personalised public transit service answering the direct needs of the community, while reducing maintenance and personnel cost for local governments ensuring continued economic viability and convenience. What’s not to like?