Martha is a modern elevator management system, equipped with voice interaction and face recognition modules, which learns the behavior and habits of its users, and which was awarded in the PropTech Festival 2022 competition. Its development would not have happened without cooperation within ITCORNER. Read the story about the innovative project by Vazco, QuantUp, IDEACTO, and Alfa-Net for the ZREMB company.
How did the cooperation at ITCORNER begin?
It all started with information shared in ITCORNER by Rafał Pisz from QuantUp. It was about an interesting and ambitious project for ZREMB (a Polish company with a long tradition, well-known for building lifting equipment). The idea was to revolutionize the market by introducing elevators able to engage their users by i.e. talking to them. In addition to traditional buttons – being not quite traditionally reproduced on a touch screen – users were to start and control elevators using voice commands. At the same time, the machines were to be smart enough to be able to respond accordingly.
Also, the solution was to recognize user behavior patterns, analyzed with the use of AI. In the initial phase of the technology’s development, this “smart elevator solution” was only one part of a comprehensive elevator management and maintenance system. Solutions that were added with time included admins defining the content displayed in the cabin and controllers able to standardize their work by using an app created for generating internal and external reports on the technical condition of any elevator.
Here’s a few words from an ITCORNER member, Michał Zacher, the CEO at Vazco, about how the project turned out:
- Together with QuantUp, we helped ZREMB win funding of PLN 3 million.
- The approach we proposed reduced the project budget from PLN 3 million to PLN 500,000 (a 6-fold reduction in budget). We reduced the implementation time from the estimated 2 years to six months (a 4-fold reduction in implementation time).
- We leveraged our experience in effective implementation and setting up of innovation projects. Thanks to the lean approach, we achieve our clients’ goals in the most optimal budget and time.
- We found a way to build a partnership with Wrocław University of Technology. A group of students carried out a pilot project for us to investigate an innovative approach to speech recognition mechanisms in the Polish language.
- We are not specialists in AI/ML, and for the ZREMB project this was an extremely important component. In this area, we joined forces with QuantUp.
- Neither are we hardware specialists. In this field, in the initial stages of the project, we took advice from Alfa-Net. It’s through our joint activities in ITCORNER that we met its owner, Andrzej Sobański.
- For UI/UX preparation, we benefited from the support of ITCORNER member: IDEACTO. Thanks to our joint activities in ITCORNER, we had built mutual trust, so the choice of a business partner was obvious to us.
- It was a pleasure to cooperate with such a professional group.
Rafał Pisz, CEO at QuantUp, on how the project started and what was important about it:
- My choice of Vazco as a suitable project partner, was based on the trust in Michał built up during our many years of working together in ITCORNER’s board of directors. As an old maxim goes: “A company is like its owner.”
- ZREMB’s decision to focus efforts on fast and efficient delivery of the MVP made us very happy, as it was in line with QuantUp’s DNA. After all, QuantUp Thinking, our proprietary methodology for AI project implementation, focuses exactly on these aspects.
- The fact that I could once again work together with Andrzej, Nikodem and Michał was a reward in itself.
What did the client expect?
- A fast delivery – a rapid response MVP prototype that was to be further analyzed by the client.
- An interdisciplinary approach – given that industrial automation processes were to be activated and controlled via web-based solutions.
- Opportunities for further development – a product created with its potential future capabilities in mind, which was ultimately to be a comprehensive, multi-layered solution for elevator systems.
How was Martha created?
While working closely with the client, the general idea was transformed into a set of definitions. By identifying the key technical and business objectives, the purpose of the project was identified by describing key functionalities and visualizing them on sketches. As a long-term investment, the project had to be carefully planned in terms of financing, and Vazco was also helpful in this regard.
Vazco, together with the client, has decided to start the work from a prototype that would include all key functionalities. This meant that it would be operational within a few weeks so that the client could analyze the project in more detail and decide on its final shape. The work was divided into several interconnected sub-projects, and one of them – face recognition and behavioral patterns – was handed over to QuantUp. However, the rest, in particular the voice assistant module and the heart of the entire solution, API Gateway – a Node-based server that mediates between all the components of the system and the elevator’s industrial controllers, which it accesses via the CANopen protocol – was taken over by Vazco
What results were achieved?
By providing a prototype, Vazco has laid a solid foundation for the final product. A developed solution is easy to maintain and expand. All key functionalities – an internet-based elevator control solution, voice interaction, face recognition, and support for elevator maintenance and management – are already available. Technology choices have been thoroughly researched and implemented: Node.js for the Gateway API, React/Electron for the cabin app, and Meteor/React for the management and maintenance system.
Martha won the PropTech Festival 2022 competition, alongside ParkCash and Żabka Nano.
Why is it important?
It is thanks to the cooperation in ITCORNER that a strong network of contacts has been built – owners of companies who trust each other and know that they can deliver high quality. None of the above-mentioned entities would have been able to carry out the project for ZREMB on their own. It would have been very time-consuming to search for such companies on the market for this innovative project, and without any guarantee of finding such proper competence configuration as we had in place in this case.
It was ITCORNER’s environment that allowed Vazco, QuantUp, IDEACTO, and Alfa-Net to build what Charlie Munger once called a seamless web of deserved trust. It is this type of network that can build Poland’s innovation and allow us to create interesting and ambitious things.
In conclusion, it is thanks to this synergy – based on business trust – that an innovative, at least on a Polish scale, project has been created.
Paweł Tymowski, the CEO of ZREMB Sp. z o.o., which is the originator of Martha, sums up the process of cooperation with the mentioned companies:
- In the process of developing Martha, we needed twofold technology.
- On the one hand – industrial technology, in terms of designing controllers and modern elevator systems heading towards regenerative solutions (and in line with the circular economy). This we had in place – it is our domain.
- On the other hand, there was the “digital” technology – designing the backend and frontend for the interface and creating a model based on AI algorithms – which was to be responsible for the “soul” of Martha. We had no competence in this, it was not our business domain.
- Support from a “consortium” of ITCORNER companies was important for us because everything was done under one umbrella. The team comprised of these companies was so complementary that we were able to implement each of the “digital” stages without worrying that we would have to look for another partner. This was beneficial for us both in terms of time and financially.
