Batteries are everywhere: from handheld consumer devices to grid storage and electric vehicles (EVs). They’re helping manage the hundreds of thousands of TeraWatt hours (TWh) worth of energy consumed globally every year.

But with use comes wear….as a battery ages the available energy it can store drops. Imagine a petrol car with a 60L tank that reduces to 50L, then 40L, then 20L … You’re filling up but not getting as far. Welcome, to EV range anxiety.

Unfortunately, monitoring this drop in tank size over time is difficult. In an EV you could drive a set course every day in perfect conditions with perfect weather until you run your battery out. But what about when you don’t want to deplete the battery? Like in a backup power environment?

The University of Waikato’s Prof. Jonathan Scott has developed a new tool designed to predict the changing capacity of a battery over time, and thus also the effective remaining lifetime of a battery. The new method is much more sensitive than existing techniques. Our technology is underpinned by new approaches to battery testing no one else has achieved before, coupled with new models of battery behaviour no one else has previously discovered.

We’re currently commercialising this under the brand eisgage [pronounced ice-gauge]

We’re partnering with battery makers, power component suppliers, and end users.

For more information, contact Commercialisation Specialist Anna Henning at ahenning@waikatolink.co.nz

In addition to Chronoptics’ successes in automotive applications and other markets, as we’ve mentioned earlier, the University of Waikato time-of-flight research team continues to develop new insights and techniques to improve the technology. Their recent research is opening up opportunities in zone-based smart-sensing for safety, security and automated touchless interactions at ranges of up to 10 metres.

Existing distance-sensitive technologies in these application areas (like ultrasonic, radar or infra-red sensors, for example) have no, or very poor spatial resolution. An infra-red sensor can detect if there is movement in a room, for example, but it cannot say where the movement is in the room. Active ultrasonic or radar methods can locate objects in 3D space, but they do so by scanning a beam across that space, which means the location and motion of an object is measured very imprecisely and slowly.

In contrast, time-of-flight methods measure the distance to objects in the image at every pixel simultaneously, and precisely. Recent improvements in the technology mean object motion and reflective surfaces in the field of view, which had previously made safety, security and touchless interaction applications very challenging are now much less important than they were previously. WaikatoLink and Chronoptics are currently exploring a range of opportunities in these areas, from easily movable virtual security and safety fences to interactive digital billboards.

If you may have an interest in applications in this area, please contact Commercialisation Specialist Doug Hillyer at doug@waikatolink.co.nz

We mentioned previously that WaikatoLink and Parkable Ltd had partnered to develop next generation parking management technology, using machine learning and computer vision on the edge. With investment from KiwiNet and the University of Waikato, WaikatoLink is developing a computer vision model for processing video images, which will be integrated into Parkable’s existing car park management software.

Since then, the development programme has gone very well and we expect to have a working pilot system in operation in a NZ carpark by the end of May 2021. Carpark owners are increasingly finding benefits from working with Parkable to improve carpark utilisation, while car owners are finding it easier to find a place to park. We expect this University of Waikato technology will help Parkable provide even more benefits to those groups. You can find out more about Parkable at https://parkable.com/

For more information, contact Commercialisation Specialist Shlok Kant at shlok@waikatolink.co.nz

We’re excited to announce that WaikatoLink and Parkable have partnered to develop next generation parking management tech using machine learning and computer vision on the edge. WaikatoLink is developing a computer vision model for processing video images, which will be integrated into Parkable’s existing car park management software to record occupancy.

What does that mean in practice? With vision software, you could implement a camera that senses the presence of vehicles, tracking how many parking spots are available in real time. 

This is a technological solution aimed at small and medium municipal and private car park owners, with the potential to dramatically change the price point for car park management. Instead of existing in-ground technology (e.g. parking sensors), the new system will process video data in real-time, monitor multiple parking bays at once, and run on several inexpensive computers, removing the need for high bandwidth, expensive servers and cloud processing. 

This technology is being developed by researchers from University of Waikato’s machine learning group, using world-leading methods and technologies that enable machine-vision and machine-learning algorithms to operate in real-time on low-cost ‘edge’ hardware.

For further information please contact Shlok Kant shlok@waikatolink.co.nz or Matt McMahon mmcmahon@waikatolink.co.nz

The Weka Workbench

The Weka Workbench is among the most popular machine learning frameworks in the world. It was developed by computer scientists at the University of Waikato and is used by entry-level data scientists and large multinationals alike, in applications from business intelligence to robotics. It contains a collection of data pre-processing tools and machine learning algorithms wrapped up in an easy-to-use graphical interface.

Weka is a great example of commercial open source software. By choosing a commercial open source model, WaikatoLink has made the software freely available to non-commercial users while generating revenue from issuing paid licenses to commercial users. This revenue helps to support research and continues to improve the platform.

Find out more about Weka here: https://www.cs.waikato.ac.nz/ml/weka/

Evironmental scientists and monitoring agencies analyse water samples to understand our climate, ecology, and the impact of industrialisation and human action on our environment. However, our understanding is limited by our ability to gather frequent and high-quality water samples, which are costly and time-consuming to collect.

The University of Waikato has developed the SYP Automatic Fluid Sampler, purpose-built for the requirements of field-active climate scientists, environmental researchers and water monitoring agencies. University researcher, Dr Adam Hartland, is using the device to build models that can better predict weather patterns and climate conditions. The device allows Adam and his team to conduct rigorous research in new, remote areas.

The prototype unit is in field testing and delivers:

• Reliable collection and storage of up to 58 discrete, silicon-sealed samples
• Up to 12 months of continuous operation using AA batteries
• Gravity-fed or pumped sample collection
• Concurrent data logging of temperature, pressure, humidity, full vial events
• Easy transport by disassembly into two halves. Custom carry-bags available.
• Easy set-up and configuration using an intuitive smartphone app
• Highly programmable sampling schedules based on time or sensor data

The device will be ready for manufacture and sale in Q4 2020. For pre-orders and further information contact Doug Hillyer dhillyer@waikatolink.co.nz or Matt McMahon mmcmahon@waikatolink.co.nz