Movement Analytics for Assets and Resources

A challenge for industry is determining what method should be adopted to transform tracking data into valuable information for operations management and strategic insights. While the Internet of Things (IoT) and other object tracking are well recognised, significant limitations exist around situational context and intent, incorporating physical constraints, trajectory discovery and quality control and detection.

There is scope to integrate machine learning and other methods to sense location and activity more seamlessly in complex factory processes, which in turn will add immense value within industrial management and automated or algorithm design. CSIRO seeks to collaborate with Australian manufacturers to define needs, develop, adapt, and integrate CSIRO-developed software and smart tags into their indoor manufacturing operations.

Movement Analytics Solutions

CSIRO is developing Movement Analytics solutions to enhance safety and productivity while integrating with existing capital assets. Methods are being developed for movement and trajectory analysis with respect to computer-represented background knowledge. 

It will engage AI techniques from the areas of:

• knowledge representation
• probabilistic reasoning, and
• machine learning based time series data analytics.

Movement Analytics solutions will capitalize on existing statistical clustering and time series prediction or classification approaches. A broad range of time series data analysis approaches can be explored based on different scenarios and it will not require new fit outs, changes or re-designs of factory floor IT systems.

This cost-effective approach is backed by a straightforward and inexpensive roll-out secure intelligent digital network of low cost, low power devices that couple location and object status information with “microcamera” devices.

Enabling vision capabilities will allow greater situational awareness of a factory floor, for example it may inform whether a forklift or an autonomous vehicle is pulling a fully loaded or empty cart.

This product is envisioned as a modular starter kit for digitalisation and factory floor logistics.

A kit comprises:

• An intelligent secure IoT network with distributed machine vision
• A plant specific application with customised AI to implement movement analytics integrated with factory level information and decision support tools.

This Movement Analytics product combines two crucial features for companies starting out on an Industry 4.0 technology adoption journey.

• Out of the box IoT security and fault tolerance, with indirect integration to existing capital assets.
• Real-time factory supervisory information management and decision support.

Practical applications

A tag based IoT device network accompanied by machine vision and machine learning can transform complex factory processes, providing staff with the real time information they need for efficiency, safety and quality control.

This system goes beyond tracking and examples of out‑of‑the‑box applications include:

• The tags can be camera fitted to cost effectively monitor solutions for non-networked equipment by accessing readouts and dials.
• The tags can be networked to communicate with each other within proximity and interpret the type of ‘entity’ in their cluster. One example benefit of this is enabling autonomous detection and reporting of safety and security anomalies.

Proposed applications

• Manufacturing assembly, construction and warehousing sectors
• Fixed asset congestion avoidance and optimal temporal and path planning
• Product and production line monitoring (ability to detect product queue build up or failures), and
• Analysis of sub-assembly assets, and de-identified human resource movements.

If this technology application presents as potential for adaptation or integration to your manufacturing processes, or your business has digital integration challenges that we may be able to assist with, we would love to hear from you. Contact us via the information below.