Your submission was sent successfully! Close

Thank you for signing up for our newsletter!
In these regular emails you will find the latest updates from Canonical and upcoming events where you can meet our team.Close

Thank you for contacting our team. We will be in touch shortly.Close

  1. Blog
  2. Article

Maarten Ectors
on 13 March 2017


The world is becoming software defined and most people don’t realise what this means until software apps and app stores invade their day to day objects like elevators. This blog post is about the smartest elevator demoed at MWC17 and the future of elevators with app stores.

What happens if we add artificial intelligence to an elevator? Add an app store so anybody can write different apps for elevators? Add Alexa so elevators can talk and listen? Allow some of the smartest hardware, IoT software and cloud to redefine elevators? That is what we demoed this year at MWC.

For those that want to understand the technical hows, please scroll to the end of the article. Everybody else, just read here how smart elevators will change your life.

Saving peoples lives

What is the worst thing that can happen in your building at work? Some disgruntled ex-employee with a gun? A terrorist? A thief? The solution is to convert elevators into traps. Just like a mouse trap but instead for criminals. DataArt trained an AI model that finds out if you are wearing a ski mask [and wants to hide your identity]. The idea was to also train the model to find out if you were holding a Kalashnikov. For obvious reasons we did not want some of the smartest brains to travel with ski masks and fake Kalashnikovs from Russia to Barcelona, so you will have to imagine the gun yourself 🙂

The elevator will detect danger and will go into “criminal trap” mode. The criminal will see the doors close, the security services being called and them being trapped inside the elevator until they get arrested.

This is however only one AI model. You can easily train others. Look for somebody being 30 seconds on the ground and call for a medical emergency. Small children being alone in the elevator, can trigger another alarm. Train the model for all employees and visitors ‘ faces and you can track if people are getting out at the correct floor.

If rich and important people don’t think their security and health is enough of a reason to want a smart elevator, then they can subscribe to the Elevator VIP service. For a $1,000/month the VIP gets priority. You get into the elevator with a VIP and the elevator will kindly tell you that this elevator is reserved and will go immediately to the Penthouse and not stop at your floor. The VIP can even choose the music or other entertainment options they want.

The next time you go to any department store, elevators can be listening and talking to you. The Alexa demo allowed the elevator to be voice controlled. The bigger thing however is that VoiceWords can be enabled. VoiceWords are like AdWords but spoken. Any company is able to bid for the spoken words that are linked to the floor the elevator will go. See the demo video for an example but think: “On the third floor, you can’t miss our special offer in the department for…”. Combine Alexa with the camera and you can now change the messaging depending on who enters the elevator. Older men in business suits will get a different message from young teenagers getting into the same elevator. Finally you can also add time based intelligence on to the elevator. During lunch hour, you get publicity about discounts if you go and eat at a new restaurant two blocks to the left.

Finally you can add the “boring” aspects of predictive maintenance and make sure the elevator is serviced before it breaks. Given that you now have a revenue generating elevator, the services company will want to fix it before it breaks. By opening an app store on each elevator, the apps each building buys might be different. Giving a uniqueness to each building as well as an element of surprise. Are you sure you want to take the elevator during Halloween!!!

By changing the economical model, elevators go from cost centres to revenue generators. Bending steel no longer is the key differentiator. So all the costs of creating the elevator should be lowered because it stands between you getting app store revenues! How to do this? Open source the design of the elevator. Let China cost optimise the elevator. As long as you control the app ecosystem on top, your enterprise can make money. In the future you can even go to a model whereby the elevator is given away for free in exchange for a large part of the app revenues. Imagine free elevators that capture criminals, save people’s lives, entertain you during Halloween and much more!

What do you need if you want a smart elevator?

The artificial intelligence models were created by DataArt as well as the Alexa demo. DataArt has lots of experience in edge and IoT computing. Check out their open source IoT platform if you want to start building something yourself: DeviceHive. Automatic calls were made via RestComm from Telestax.

To train the models you can use Google’s Tensorflow on top of Kubernetes, trained on GPU farms. In order for your team to get started easily we are working on easy to use Deep Learning infrastructure automation via Juju Charms. Contact our cloud team if you need more information.

Charlie Isaacs, the CTO of IoT of Salesforce, surprised visitors at MWC on Monday and Tuesday by personally demoing how the Salesforce IoT Cloud automates your workforce in case your smart elevator needs anything.

To do the processing of the videos coming from the elevator cameras you can use Dell’s IoT Edge Gateways. Especially the high powered Intel i7 will be needed for complex edge processing.

Controlling the elevator is best done via App Logic Controllers or ALCs. ALCs are revolutionising the industrial PLC market because they are exponentially cheaper, faster and easier to programme. Great examples are KunbusRevolution Pi and UniPi‘s Neutron.

Controlling complex predictive maintenance and other edge IoT use cases can also be done with Azeti‘s and CloudPlugs ‘ IoT Edge platforms.

Want to write your own apps for elevators? Be sure to check out Snapcraft.io, tutorials.ubuntu.com and ubuntu.com/iot.

Industrial IoT needs App Stores and a vivid partner ecosystem. Check out ubuntu.com/core and contact us.

Related posts


Karen Horovitz
18 March 2024

Canonical accelerates AI Application Development with NVIDIA AI Enterprise

Kubernetes Article

Charmed Kubernetes support comes to NVIDIA AI Enterprise Canonical’s Charmed Kubernetes is now supported on NVIDIA AI Enterprise 5.0. Organisations using Kubernetes deployments on Ubuntu can look forward to a seamless licensing migration to the latest release of the NVIDIA AI Enterprise software platform providing developers the latest AI ...


Yalton Ruiz
9 November 2023

Canonical Kubernetes enhances AI/ML development capabilities with NVIDIA integrations

Kubernetes Article

In recent years, Artificial Intelligence (AI) and Machine Learning (ML) have surged in importance. This rise can be attributed to a massive influx of data, enhanced computational capabilities, and significant advancements in algorithms. These changes have empowered various industries and society at large, resulting in cost-effective produ ...


Felicia Jia
16 April 2024

Canonical presence at Qualcomm DX Summit @Hannover Messe

Internet of Things Article

At the world’s leading industrial trade fair, companies from the mechanical engineering, electrical engineering and digital industries as well as the energy sector will come together to present solutions for a high-performance, but also sustainable industry at Hannover Messe. This year, Qualcomm brought its DX Summit to Hannover Messe, pu ...