The shift from cloud-to-edge

Debom Ghosh
3 min readDec 13, 2021


It is 2025, you are riding a autonomous car. You see a pedestrian crossing the road. Would you allow your car to rely solely on the remote servers to take such a time-critical decision? Imagine that in this case your car detects the intrusion, sends the data to a server, in return gets a response and then takes a decision. This architecture although being very fast, is questionable under critical situations. If there are sensors that know the “intrusion”, why can’t the car take its own decisions?

Let’s take another situation, suppose you are giving instructions to your local home automation devices like thermostats, your IoT Television, the vacuum cleaner or the smart curtains. Would you be fine with the public cloud eavesdropping on your local devices? In other words, the cloud becomes a part of your private life — when you wake up, when you feel hot or cold, your TV program choices, your conversations at home! The problem is more prominent for your IoT devices. By the next 5 years, an anticipated 20+ billion devices are likely to be connected to the internet, creating an incredible amount of data every minute. With these challenges, storing, maintaining and working with this amount of data is a an uphill task for any IoT enterprise. Jittery networks, low communication bandwidth, response times of different APIs, insecure gateways to the cloud and server.

What if there were options to make these devices self-sufficient, allowing them to take their decisions locally “on the devices”? This is what making them “Edge-capable” means. The solution is processing your IoT device data locally rather than sending it to the server, enabling internal communication in between the nodes and sensors. This reduces the latency, enhances security, and results in more robust and scalable application performance.

Edge Computing is shifting the infrastructure closer to the end user, bringing storage, maintenance and decision making to the devices. The shift from Cloud to Edge is already a reality, and at Seashell we believe it is crucial for realizing the Internet-of-things’ potential. And here are its benefits:

  1. Privacy — Your commands to Alexa/ Google devices or maybe Jarvis, stays inside the devices. The call-to-action is taken by the device itself instead of sending it through gateways to servers or clouds.
  2. Real-time responsiveness — There is practically no latency as Edge Computing enables local processing of data from the sensors present in the device, so that things like cars, drones and robots can take decisions at the split of a second.
  3. Reliability — Another benefit of Edge-enabled IoT devices is that you reduce the chance of the device failure considerably in case of remote server breakdown. Since these devices are meant to work off-the-shelf, being able to function remotely at all times.
  4. Data transmission costs — Data transmission is getting more and more expensive. With Edge capabilities, the data is stored and processed locally, drastically reducing the data volume which is transferred through the cloud.

We at Seashell help companies leverage this benefits on their IoT applications. By providing users with several battle-hardened open source components like Docker, Nomad, Consul, Drago and Wireguard we enable a cloud-native development experience also at the Edge!



Debom Ghosh

A Product Manager working in the field of IoT, Edge Computing and Machine Learning topics