Our client, one of Africa’s largest energy companies, has over 50 million individual customers. Late meter submissions, slow verification, and manual input errors kept adding manual checks.
Computools created MeterSnap, a mobile app powered by neural networks that recognizes meter data and allows readings to be sent directly to the utility. For large utilities, energy software development services help connect meter submissions, operational validation, and back-office workflows into one reliable data flow.
Late submissions fell from 25% to 10%, manual corrections and onsite inspections shrank by 82%, and the data became harder to tamper with.
Our client is one of Africa’s largest energy companies and a producer of low-carbon electricity, with over 50 million individual customers. For many years, it has supplied electricity to private households and local businesses.
The company needed electric power utility software development that could support meter data collection at scale, reduce late submissions, and make readings easier to verify.
A quarter of meter submissions did not arrive on time. Verifying submitted data took extensive time, and manual input errors led to further manual checks. A vast network of outdated analogue meters and meter reading fraud added to the operational losses.
The client needed a way to collect meter readings faster, validate submissions, and reduce manual corrections. The project also required data engineering to connect meter images, customer account data, geolocation, and submission records into a reliable validation and processing flow.
Computools developed MeterSnap, a mobile app that uses image processing to read the meter and link it to the customer account, so readings can be sent directly to the utility. Geolocation validates each submission, the app keeps working with poor internet connectivity, and roles and permissions limit who can change data.
The solution used neural-network-based recognition to reduce manual input and make meter data harder to alter. Computools applied AI development expertise to support image recognition, data validation, and automated reading submissions.
The application improved how the utility collects, verifies, and processes meter data.
Computools was chosen for this project because of its digital transformation expertise and experience developing mobile apps for the energy industry. The team also has a strong track record of delivering.
Computools has a team of experienced software engineers with expertise in various technologies, including neural networks, mobile development and web development.
In this case, the design tasks encompass a range of aspects associated with developing and designing the MeterSnap application. Creating an intuitive and aesthetically pleasing user interface for the mobile app while ensuring an excellent user experience involved designing user-friendly navigation, clear control elements, and an efficient screen structure to facilitate inspectors' work and enhance customer interaction with the application.
Creating a detailed profile of the typical application user, considering their needs and preferences.
Outlining the hierarchical structure of the application's pages and navigation flow.
Drafting basic, low-fidelity visual representations of the application's layout and content placement.
Developing the final, high-fidelity graphical user interface, including design, icons, and interaction elements.
NATIVE LIBRARY (C/C++, STL, NEURAL NETWORKS)
The cross-platform library was designed for image processing and recognition across desktop and mobile platforms. OpenCV was chosen for the input frame pre-processing, contour analysis, and recognition area setting. Image recognition was implemented using a neural network-based classifier.
DESKTOP CLIENT/TRAINER (C++, QT 5.X, QML)
Our team developed a desktop application using a Qt 5.x framework and Qt QML technology. This application was used for neural network training, control, and testing. Instead of a mobile phone, a webcam was used for data input.
ANDROID CLIENT (JAVA, ANDROID SDK, ANDROID NDK, REALM DB, RXJAVA, RETROFIT2)
Our developers implemented the mobile application with native library features for image recognition, domain logic for data analysis, and auxiliary functional units. The local RealDB database, a high-performance NoSQL database for mobile platforms, was implemented to ensure the application's offline operation.
WEB API SERVER (JAVA, SPRING FRAMEWORK, MONGODB)
Centralised data storage and mobile client request processing were implemented in a web server based on Java and Spring Framework. Our software engineers settled on MongoDB as a server database due to its high performance and speed parameters.
We chose the Scrum project management approach with two-week sprints to ensure continuous delivery and accurate fulfilment of the development plans. Agile methodology necessitated close cooperation with the client to introduce new ideas and approve adjustments. Every sprint was finalised with the testing and demonstration of the implemented features.
We are impressed with Computools’ expertise in AI and mobile development. The MeterSnap application has transformed our meter reading operations, improved efficiency, and enhanced customer service. We are confident that this solution will play a key role in the future of the African utility market.
As part of Bombardier's digital transformation and IT automation initiatives, Computools developed a secure document version control system built on a distributed architecture. The system gives aircraft technicians access to the correct maintenance instructions in seconds, reducing time spent searching for documents or contacting technical support. It also streamlines communication between pilots and technical teams by minimizing requests for the correct versions of operational manuals.
Our client, a major European automotive brand, required assistance developing situational awareness and obstacle detection software. The team at Computools helped build the tool used to generate labelled data from the video stream and advised on selecting image processing algorithms, such as SIFT. The solution can detect obstacles, recognise traffic signs, create situational awareness and thus contribute to self-driving decisions. The digitally transformed system was used to train the machine learning models.
A major African utility needed a faster and more reliable way to collect electricity meter readings. Computools developed MeterSnap, a neural-network-powered mobile app that recognizes meter data, links readings to customer accounts, validates submissions with geolocation, and sends data directly to the utility. Late meter submissions decreased from 25% to 10%, while manual corrections and excessive onsite inspections were reduced by 82%.
Computools | 341 Raven Cir, Camden Wyoming, Delaware 19934, US
