July 5, 2022


News and Update

Problem: Sensible instruments for audio evaluation in IoT networks

Ooredoo invitations proposals from Startups, SMEs and Enterprises on applied sciences that present Sensible Engines for Audio Analytics in IoT networks. Ooredoo is on the lookout for revolutionary {hardware} and software program options to handle the rising want for near-real-time audio knowledge in IoT networks, enabling measurement, predictive analytics, and upkeep.

Picture credit score: goingai qua Pixabayfree license

Audio monitoring can passively present numerous knowledge units for machine monitoring, for instance, for metrics, single fault detection, and audio sample recognition. When mixed with Synthetic intelligence and Machine Studying, audio methods present nice potential for implementing proactive monitoring, predictive upkeep, and real-time intervention.

The precept is to seize audio indicators from the atmosphere or an object and analyze the sound patterns (e.g. noise from a working engine). In contrast with ‘repeatedly repeated’ patterns or by way of an adaptive studying strategy, biases or anomalies (e.g. mechanical bearing noise) could be recognized, and thus could be launched. make suggestions, warnings or actions.

One goal use case is for upkeep and repair tools similar to automobiles, building equipment, or elevators in buildings. That is usually carried out primarily based on frequent on-site service visits and time-based utilization assumptions (like hours of operation). Upkeep primarily based on remotely capturing tools well being or utilization data by way of IoT applied sciences can significantly enhance service effectivity. Any breakdowns, abnormalities, or utilization circumstances are monitored remotely and allow decreased intervention prices and proactive service planning.

Ooredoo is excited by figuring out Clever Audio Analytics options for use in varied utility eventualities associated to tools service and upkeep.

See also  After layoffs, crypto startups face a ‘harmful second’

The system is predicted to have the ability to establish defects or different acoustic situation issues and ideally find the issue. This distant system could be retrofitted to present tools and new methods. Sensible audio methods should acquire knowledge in sampling mode and/or steady operation and carry out evaluation with acceptable latency within the vary of minutes, relying on the tip utility. As well as, the system should detect and analyze audio patterns (i.e., the evolution of sound over time), which develop spontaneously over longer time scales.

Submissions for this Problem should be obtained by September 6, 2022.

Supply: NineSights