Topic: How to Build an Embedded Wallet For Ethereum (Read 210 times)

You are right, for many IoT applications it is not necessary to have a currency, we will indeed support even "purely blockchain" solutions like Hyperledger.

The advantage of having a natural and secure way of exchanging currency between physical devices should not be underestimated though. It is something far from trivial with a normal IoT device and can enable many interesting use cases. For instance instead of paying in advance for the shipping cost of an item bought online, you could put the money into a smart contract and automatically pay when the goods arrive together with an IoT sensor which monitored the shipping. The smart contract could specify that if certain conditions are not satisfied (for instance temperature over a certain range, time delays, shocks, too much humidity) a discount or complete refund should immediately apply.

Not sure if it fits your use case, but it seems you need a blockchain, not a currency. I'm watching this quite closely https://github.com/tendermint/tendermint (not affiliated in any way).

I think these guys take things to the next level and create an app framework (PoS based, so there is no mining, which is good for  IoT) to implement blockchain based  dAPPs.
They also, of course, got excited buy this ICO mania and preparing a cryptocurrency based on this, but I think this is not the main point.

If I would decided one day diving into all this crypto thingy (as a dev), that would be a blockchain framework, not another cryptocurrency. All things imho of course...

You're correct that Boron (coming from Particle, https://www.particle.io/mesh/buy/boron-lte/) is a secure device given that it's supported ARM TrustZone. But IoT security model is far from being complete just by just having that option on your IoT device. Identity and being able to verify the authenticity of the data coming from a particular device is another dimension of security model. These two come "for free" once you're using blockchain on your device as Root of Trust.

Furthermore, it's not just the data on the device which needs to be secure. Data uploaded to a centralized service are inherently unsafe since you're trusting that third party. Decentralized approach (blockchain + IPFS) offers a ticket out of data silos like Amazon, Alibaba, AT&T M2X, IBM Bluemix, Microsoft Azure, Google Cloud, mBed Cloud, and other private clouds and lets device owners to own their own data. We explain this approach in more details in our whitepaper which you can find here: http://www.anyledger.io/.

For example, the used Boron: LTE has a built-in protective module ARM TrustZone  CryptoCell-310, its protection in my opinion is sufficient for domestic use, why do you need a blockchain?

Just to clear a potential misunderstanding. The security cam would run only a wallet and not the full or light blockchain node. This is something what was explained in the blog post as a common misconception in the industry when it comes to combing blockchain and IoT. Given that the security cam has the wallet locally, it will own its private key as well, correct. In the example from the blog post, the micro-controller unit (MCU) being used on Particle Photon is belonging to similar family of MCUs like Trezor (STM32F205) which actually has a large hardware-based attack surface. Check out this great article on that topic: https://blog.gridplus.io/hardware-wallet-vulnerabilities-f20688361b88. The idea behind the blog post was to explore the possibilities and potential use cases of running a wallet on a embedded device. Now, making the device itself more secure requires much more effort of course. This is exactly what is our startup (called AnyLedger, http://anyledger.io) working on atm. We are building a highly secure, embedded wallet solution which will expose IoT devices to the blockchain ecosystem. We will be using a new type of MCUs with high security standards (ARM TrustZone) and enclose the MCU in a unibody with special welding procedure to provide temper-proof capabilities. There are more security measure we plan to add to it. You can check our whitepaper for more details.




We're definitely not implying that the every single IoT device out there should be a wallet as well. What we're trying to convey is that having connected capability for the IoT device will become commodity in the near future. Once that happens, whole new set of use cases will open up. These connected devices don't necessary have to exchange value between each other. More interesting scenarios involve their interaction with smart contracts. Think about having an audit trail for you home appliances which prove the usage time of a specific device. This can be used to determine a more fair second-hand price for example, or suggest for the potential repairs in order to extend the lifetime of of the device. We might be able to fight against planned obsolescence by requiring manufactures to enable these kind of scenarios. Not all IoT devices are home appliances btw. IoT refers to a much broader spectrum of industrial machinery, electronic appliances, wireless sensors, cars, transport systems, wearables, smart meters or any other everyday object with enhanced digital capabilities.





I think your comment is spot on this one. Currency based vocabulary might be causing a certain level of confusion. Our product is quite technical and we're still trying to find the correct way of telling its story to our community and potential customers. Given that this is relatively new application area for blockchain, we thought that "embedded wallet" term would push the right buttons in people's head. The concept of wallet is wide-spread technical term in blockchain world and it's not specific to Ethereum. Also, our startup is trying to abstraction layer for the wallet itself, so it would easy for a IoT device and its owner to move between different chains.

That is an excellent question. The blog post from above was really a proof of concept and to investigate what is possible to deliver on an embedded platform. Turned out that you can run a full blown wallet on a small footprint device. Coming back to your question. We have chosen Ethereum for our prototype, but we won't be tying ourselves to it (even though there are some promising developments on that chain like PoC and Plasma). One of the core directions for the startup we're building (called AnyLedger, http://anyledger.io) is being agnostic for the underlying blockchain. Our second iteration of the prototype will involve new hardware setup (smaller size, smaller footprint) and support for other chains out there (Stellar, IOTA or even tech from Layer 2 like Lightning Network, Plasma) which are more suitable for IoT scenarios. Given that the landscape of blockchain ecosystem is always evolving, we don't think there will be a single winner for the IoT scenarios. Therefore, we believe that having a platform which can expose ioT device to multiple chains would be a compelling offer.

Thanks for the feedback btw and if you're more interested into this topic, please check our website (http://anyledger.io). We're giving away first batch of our prototypes to the community once they are ready.

Reading the article, got some questions.


How could this work? Imagine someone has a blockchain in his security cam, which reports the data to central server and signs it so that it is unfalsifiable (if I correctly understood the use case). Now, this small device should have it's private key in it, stored somehow locally, correct?. What is the security measure if it?


Why a toaster needs a wallet? No, seriously, I can understand that blockchain, probably, has application to IoT. But why things should be built around "wallets" and "balances" there?


Why would it be necessary, funds in a wallet of my door bell? Wouldn't distributed, blockchain based consensus work here, but not involving currency-ish idioms?


Sorry for some critics, but you asked for comments. I do think blockchain can be used in IoT and such a projects have great potential, but I can't get why it's still being built around "wallets", "funds" and "balances". Is this because there is no technical basis for something different and Ethereum dominates the area?

Ethereum and IoT? Have you taken a look at the transaction capacity of ethereum? Are you sure you are in the right chain?

Hi guys,

I'm building a startup in Berlin applying blockchain tech to the IoT (Internet of Things). Here's a small demo that we did a couple of months ago, hopefully you will find it interesting:

https://medium.com/coinmonks/building-embedded-wallet-for-ethereum-c9777c1f4697

(it's a Medium article written by my cofounder)

I'm looking for feedback, every comment is welcome!