Skip to content
SensiML
  • Solutions
    • Industrial
    • Commercial
    • Agricultural
    • Wearables
  • Resources
  • Company
  • Client Login
  • Contact
QuickLogic
SensiML Joins Quicklogic To Launch a Far Edge Artificial Intelligence Initiative on May 4

SMART SENSORS: VITAL TO HIGH PERFORMANCE IoT NETWORKS

Going far beyond traditional simple switches or value/threshold sensors, smart sensors provide much richer insight using the complex high-fidelity signals originating from accelerometer, gyroscopic, electrode, vibration, thermal imaging, optical, audio, and pressure / strain sensor inputs.  But they do not stop there as these high data rate raw signals would otherwise quickly overwhelm downstream networks with massive amounts of redundant or useless data.

Instead, smart sensors locally process their physical raw sensor signals using techniques such as pattern recognition and machine learning to provide only information of value to upstream nodes in the network.  Their role can often make or break  the entire IoT application where bandwidth limitations, security/data privacy, latency, local real-time response, battery life, and wireless endpoint considerations are key factors.

SensiML TRANSFORMS HOW SMART SENSOR ALGORITHMS GET BUILT

Creating smart sensors that run efficiently and provide accurate meaningful insights can provide substantial IoT product benefits but is highly challenging in terms of required time, cost, and team skillsets.  The time has come for a smart development toolkit that breaks down these barriers allowing all device builders big and small to truly innovate with smart sensors without the burden and constraints of hand-coding algorithms.

Read more»

Today, creation of smart algorithms is largely the realm of large device OEMs who can justify the significant investment in firmware development, digital signal processing, data science, application programming and domain expertise.  Even then, these large OEM teams must limit algorithm scope to fixed-function, must-have features as the process is neither well automated nor scalable.  As examples, consider the fixed 'wake word' required to trigger smart home hubs or the fixed 'step' or 'stair climb' motion detection algorithms of consumer fitness trackers that ignore or misrepresent all other valid modes of exercise activity.

For the rest, namely mid/low volume applications and smaller product teams, without practical means for generating smart sensor algorithms at all, they are left with either cloud processing or companion smartphone applications as the only practical options despite the shortcomings these dependencies inject in application performance, latency, and user experience.

SensiML's automated learning algorithm toolkit addresses these challenges head on solving the time consuming and expertise demanding aspects of algorithm design.

For large teams, this translates to easily extensible datasets and expanded use cases allowing teams to not only offer far more functionality in the same product development timeline, but also opens up new business models where smart devices can become the vehicle for ongoing application services powered by learning and updateable sensor algorithms.

For smaller teams, SensiML's toolkit offers the freedom to build true smart sensors without over-reliance on smartphone processing or cloud analytics that constrain your user requirements.  Smaller teams are empowered to create complete algorithms with focus on their specific domain expertise and only average programming skills.

SensiML Analytics Toolkit: The solution to building smart sensor algorithms quickly and easily

Algorithm Creation at the Speed of Capture

If you can capture what you seek to detect, SensiML's machine learning optimizer can automatically generate a predictive algorithm. Inordinate time and energy spent in iterative coding, testing, and re-coding can now be focused on extending algorithm capabilities making your smart device truly smart.

Sensor Analytics At The Extreme Edge

SensiML creates optimized algorithms that execute locally on the embedded sensor node, not a gateway or in the cloud.  Thus, your applications benefit from true real-time processing where the sensing take place.

Rapid Prototyping and Proof-of-Concept Validation

With tools that address the entire workflow for data-driven algorithm design, SensiML Analytics Studio can streamline all aspects including initial prototyping and proof-of-concept work. Our tools allow developers to quickly test sensor applications on hardware in a matter of days or weeks rather than months.

Smarter Sensing Through Continuous Learning

Get your devices out of the lab and into the hands of customers sooner with ongoing learning mechanisms that allow you expand model sophistication over time through crowd-sourced data feedback and regular updates. Your users can enjoy more engaging products as they contribute to solving their own corner case exceptions.

SensiML ANALYTICS TOOLKIT

An end-to-end software suite that provides developers a straightforward process for developing pattern matching sensor algorithms using machine learning technology. Each component handles specific steps in the process to progress from initial raw sensor data collection using your prototype hardware, to automated signal processing and feature engineering, to embedded algorithm optimization, validation and testing, and post-ship algorithm updates and continuous learning enhancements.

SensiML Data Capture Lab

A client tool available on Windows PC and Android mobile platforms, SensiML DCL enables convenient and efficient data collection, data cleansing, labeling, and metadata annotation of your sensor application datasets.

SensiML Analytics Studio

A server-based ML data compiler that uses your sensor data to deliver device-optimized algorithm code that runs on low-cost sensor MCUs.  SensiML Analytics Sutdio takes the guesswork, time, and risk out of your projects.  Pick your target device, your required accuracy, representative data, and let SensiML Analytics Studio figure out...

SensiML Test App

Once your embedded code has been generated, SensiML Test App can be used to quickly and efficiently validate the proper behavior, accuracy, and performance of your algorithm empirically on actual sensor hardware.

SensiML API

Providing a straightforward interface to utilize and extend your SensiML algorithms, our API also provides the means to channel new data discoveries from users back to your modeling pipeline for continuous learning and functional improvement for your devices.

Industrial Use Cases

Agricultural Use Cases

Commercial Use Cases

Wearable Use Cases

Added Services

Beyond our standard product platform offerings, SensiML offers custom services to help developers maximize their productivity and further reduce development resource needs and risk.

Enterprise Level Support

One-on-one classroom training, direct technical support from SensiML developers, weekly technical project support calls

Custom Engineering Services

Turnkey algorithms with customer supplied data, Consultation services (test design, data, collection integration)

Custom Development

Custom product feature requests, Extension to non-supported embedded platforms

Contact Us

About Us

SensiML

SensiML (say "sen-seh-mul") is a spin out from Intel Corporation offering cutting-edge software enabling IoT developers to quickly and easily generate app-specific pattern recognition code that transforms rich connected sensors into smart actionable event detectors.

More Info

News / Events

QuickLogic

SensiML joins Quicklogic and other ecosystem partners to launch a far edge Artificial Intelligence (AI) Initiative with a webcast on May 4, 2018.

More Info
SG2018Logo

SensiML named a Top 10 company at Startup Grind Global Conference 2018

More Info

  • Privacy Terms and Legal Notices
 Copyright © 2018 SensiML Corporation