It is claimed that Haven is here to set a new standard for robotics.
What makes it different from Agents?
So, how does HAVEN exactly work?
Why would anyone use HAVEN?
What’s the main incentive behind partaking in the ecosystem?
Let’s start off the most important, which is the difference between a robot and an agent.
A Robot is a physical machine that can sense, act, and decide in the real world.
An agent is a software (or conceptual) entity that perceives and acts to achieve goals — physical or digital.
Haven provides a virtual on-chain simulation of a physical ecosystem, albeit simple at first, with infinite possibilities. The utmost obvious use-case that comes to mind is collision testing.
Imagine subject A has a cleaning robot inside his house. It’s a very expensive item, so the subject does not want to risk running it without securing that there are no collisions.
Logically, such robots are made for this scenario, but everything in the software world, can (and will) fail in a certain context.
How would subject A prevent the failure of this exact context? Simulation.
Agents can only retrieve and treat information, but they cannot simulate collisions size, physical statuses or traits, not in real time. An agent can be trained to know how to react to a problem (what function to call), but it cannot be trained to avoid the problem itself once presented. It is not a physical object, nor is it a representation of such.
Robots, on the other hand, in Haven’s case, are a virtual on-chain, immutable representation of real-life robots with their same exact functionalities and limits. An agent may be able to tell the temperature both in Russia and China at the same time — but this is something that would be impossible for a physical robot with a physical sensor, which is what makes robotics exciting, and sometimes even more ‘humanly similar’ to agents themselves, which are rather software entities.
Let’s get technical here!
Haven works with a simple object oriented framework in which every robot represents an object of type robot, whereas every other object is seen as secondary, on a separate class.
The environment variables themselves, since they’re only relevant to the robot (i.e, if the robot is inside a house, the fact that it’s raining outside does not change the simulation’s outcome), following programming principles, are not a class of its own, but included in the robot’s own data, since this is considered the main subject of the program.
What does this mean to the user?
It means that robots can be fully customized in terms of abilities, commands, sensors, and lack of. Robots can be useless. Robots can perform tasks and simulate real world scenarios that its physical form would struggle with. A physical robot collision could cost damage up to the thousands for companies — however, a virtual robot collision only means an ERROR message on a terminal, plus a pat in the back to “fix it”. Most importantly, everything happens on-chain, transparently, allowing robot owners to permanently access robot logs, so long as the chain is active, even in the rare case of HAVEN’s app being inactive.
There’s 3 main technical aspects that make HAVEN’s backend unique:
Every robot runs on its own thread, for everything. This means that a looped-robot that’s stuck with a never-ending command, will not block the tasks of another.
There are no hardcoded limits on the potential skills a robot can develop over time. It’s up to the user’s potential, not the dapp’s limitations.
Simulations are tied to the robot objects, therefore providing the chance of simulating real physical events, such as wind, pressure, rust, battery life, quality, temperature, and any imaginable and relevant (to the user) condition there is.
Haven aims to provide a key service in robotics — which is the virtual simulation of physical robots.
Majority of people are looking at agents, and they’re right; agents are clearly the future of multitasking for human beings.
However, not every task can be performed online, as much technology as the world has, plumbers, manual labor, construction, cannot be replaced by simple code.
Even though code runs behind it, there’s a part of the world — the physical part — that agents cannot ever access to.
This is why robots are going to become the worldwide norm for white collar jobs. This is a reality that’s closer than most realize, very likely less than a decade.
Physical robots can, ideally, be tested in the physical world, with the risk of damages to equipment. Of course, nothing will ever beat a real test, but not every version of a product is meant to be tested on production.
This is where HAVEN comes. Haven aims to become the main framework for processing virtual robot task simulation with physical entities and rules. It’s code, but it represents the world. It’s not just information. It’s positional data, object and sizes, weights, capabilities and lack of.
All of it always accessible, on-chain.
Rather sooner than later, not only the big corporations, but small-to-middle sized companies will have to adapt to the fact that robots have become the new norm for the future, allowing maximum profitability without the risks of human nature.
HAVEN welcomes their simulations.
If not building a physical robot or testing capabilities of one known-model, why use HAVEN then?
When entire businesses models are changing the way they assimilate problems with robotics, being part of one of the main frameworks for physical data simulation is an undeniable advantage, both in knowledge and practice. Web2 Technologies similar to HAVEN will be flooding the market very soon — because they’re needed.
Acquiring expertise on such tools could determine whether an individual keeps their job, the same way typewriters were replaced by skilled individuals with computers.
As far as an economic incentive goes, the project is providing a $500k USD acceleration fund, aimed to support top-builders of the ecosystem.
This is not all, however. The most important perk of HAVEN, as for user incentive goes, is the ability to tokenize robots and its functionalities, with the eventual potential of robots being fully on-chain, without even requiring a centralized backend at all. This is HAVEN’s final goal.
In Conclusion, what differs HAVEN from upcoming frameworks is that one of the key focus points is permissionless on-chain records of simulation results, which means that decentralization will keep logs permanently online (and free to access) without the need of corporations, or big-data users, to store the results themselves, merely having to process the outcomes instead.
There is no better place to store immutable simulation data than web3, which brings independence to users & allows unanimous transparency of results for companies, features which web2 may lack of, lately, most of the time.