A collaboration of academic and industry experts with Gaiax, a Japanese company known for its social media and application enterprises, tests a system software that introduces a city-wide lidar network.
It uses the blockchain system, and is proposed to make real-time navigational data free for every single one to use and process.
An Alternative Step Towards a Smarter City
Blockchain’s decentralized processing and storing of digital information has spawned many different ideas, often related to the indirect handling of real-world elements and assets. This, of course, includes the ever-growing application of cryptocurrencies, something the world is currently enjoying or suffering, depending on which side of the industry you are on.
Ideally, this would then create a city-wide digital picture map, which all devices, units, vehicles, and other related machines could directly use to power their autonomous navigation functions.
The initial part of the proposal calls for the participating academic and business entities to develop the lidar image sensors, which would then be deployed strategically over a huge portion of an urban center.
The aim is to create a “zero blind spot” field, where every tiny traversable corner of the region could be mapped for analysis. The resulting real-time, updating data could then be constantly fed to things like delivery bots, security drones, and most importantly, self-driving cars.
The proposal even shows applications such as earphone-based notifications for walking pedestrians.
The second part of this would attempt to develop an independent data aggregating system (this is presumably where Gaiax comes in). This is to mitigate the inefficiency of the blockchain system when dealing with real-time data processing, especially with potentially thousands connected to the same network, should this proposal become a working reality.
The Perfect Self-Driving Car Augment?
While a single vehicle equipped with its sensors could relatively be able to analyze the situation of the roads on a general scale, the fact of the matter is that, there are still many unknown variables that could put the AI’s judgment out of whack.
The infamous self-driving uber crash incident last 2019, for example, showed how an autonomous car could still fail to respond in time despite the system detecting the victim in advance.
Thus, the biggest advantage of this blockchain lidar network + data aggregation system is redundancy and information exchange. Not only do you get far better redundancy with the sensors placed independently away from the machine or vehicle using the data. But, you can also interdependently put a check on each other, so that every single thing on the lidar network knows each other’s existence, and each independent AI can act accordingly upon this information.
Then, Why Not Just Build a Dedicated Lidar Network?
Indeed, it is possible to simply create a dedicated network that would collate and store all the image sensing data for some centralized AI to process. With a robust design from one entity, or at least two closely collaborating companies, the blockchain concept would even seem like a pointless redundancy (pun intended).
However, you can probably already see the issue when the words “companies” and “centralized” are thrown into the mix.
In the end, those tiny bits of unreadable data would break the “zero blind spot” field, forcing the return of the unknown variables that plague autonomous vehicles today.
Worse, some malicious entity could just hijack the system to cause harm to those connected to the dedicated network.
This is why the proposal deeply stresses the need for the data to be freely “chopped to bits” and distributed over the entire network as publicly decentralized information. In traditional blockchain systems, there is a need for approval by every single “member” (computing unit) of an archive before confirming any modified data. The same goes for this one.
Any positional data that the city-wide lidar network produces, would have to be confirmed by the rest of the other separate sensor units and other vehicles connected to the system.
Human-Free Roads, Finally…
The result? Well, according to the joint research and Professor Ryoichi Shinkuma of the Kyoto University Design School, it should be a huge step towards Society 5.0. After the ongoing preliminary tests set in different designated parts of the country this year, they are also more likely to push forward to other related “smart city” technologies.
That being said, from our limited perspective, this might just ultimately become the endgame vision for self-driving vehicles in the far future: the idea that humans will no longer be part of the equation.
Think about it. Suppose every other autonomous car could talk to another one several tens of miles away, and be able to readjust its course to include each other’s existence in the equation. Would something like that Uber incident ever happen again?
… Err, actually yeah.
But we guess, that’s exactly why the buildings with lidars are going to check out those pedestrians for your cars to “see” and analyze.