IoT in the Animal Lab: How the Internet of Things is changing the future of preclinical research.
Today, Iot is at the top of the list of disruptive technology. According to Gartner, a leading research and advisory company, 25 billion Iot devices will be installed worldwide by 2021. In 2018 alone, more than $3.5 trillion dollars was been spent on Iot.
How does Iot fit into your animal research lab? We will introduce some of the technical, operational and financial issues to keep in mind as you consider an Iot future.
What is the Internet of Things?
Gartner defines Iot as a network of dedicated physical objects (things) that contain embedded technology to sense or interact with their internal state or the external environment. This excludes general purpose devices, such as smartphones, laptops and tablets, as they are general purpose computers designed for human interaction, not specialized devices that monitor or interact with the environment.
Iot devices are typically small sensors that communicate directly via Wi-Fi, cell signals or through a personal area networks (PAN) with specialized Internet gateways. Iot devices can also include artificial intelligence (Al) algorithms that can make decisions to send alerts, control valves or toggle switches.
Iot devices are said to live on the edge where they form the interface between the physical and cyber universes. This interface between the edge and the cyber universe is modeled after biological systems. Humans, for example, have five senses with many neurons attached to the brain.
Similarly, you can think of Iot systems as sensors connected to a computer brain. In this model, the computer networks act as neurons that transfer the information to the brain. Some sensors can respond instantly to hazards, like closing a valve if water is detected in a cage. Some simply capture data from sensors and transfer the sensor data to the cloud. Iot systems connected to large computational "cloud brains" are now at a technological stage of maturity where they can be deployed for business value.
Cases that justify the expense of purchasing and maintaining Iot systems are easy to find--however, there is still a high risk of failure if not planned and executed carefully.
Iot in animal rooms
Animal cage cleaning is a regular occurring event that is stressful for research animals. Rodents, for example, are sensitive to odors and pheromones introduced into the micro-environments when animal caretakers intrude.
Stress induced by frequent cage intrusions can also lead to aggressive behaviors in mice, decrease their body mass and increase pup mortality. Cage cleaning is also a drain on research resources and increases health risks to both the animal caretakers and the research subjects who may be exposed to infectious agents.
On the other hand, cages that are not cleaned frequently enough may accumulate high levels of ammonia, which can have adverse physical health effects on the animals.
The costs of labor, bedding and autoclaving, along with scientific and animal welfare concerns, are driving research organizations to develop data driven approaches to cage cleaning.
Iot sensors are a solution for automating cage condition monitoring and developing principled metrics for cleaning cages. Smart cage manufacturers have shown that cages equipped with sensors can detect unhealthy ammonia levels, humidity levels, temperature, light and sound.
The cost of Iot technology is dropping rapidly, and it will soon be possible to equip individual mouse cages with sensors connected to smart "things platforms" that integrate Iot data with scientific study data. With Iot prices dropping rapidly, we can expect a quick transformation from calendar-based cage changing schedule to Iot data driven scheduling.
Turning hype into business value
As you plan Iot, be careful of traps. You want good business value, and you need to deploy your Iot with technologies that can scale. Avoid creating new Iot data islands but stay agile. Agility will allow your organization to continually adopt new technology and adapt to business changes with minimal new investment.
Small Iot projects conducted in individual research labs can result in quick wins. However, to gain full value and meet IT security requirements, you need institutional support.
The first step for an organization planning to adopt Iot is to start an Iot Center of Excellence (CoE). The CoE should be comprised of IT professionals with a background in Iot technology, architecture and security.
The CoE must also include stakeholders from all areas of the business. A CoE that does not include and engage stakeholders will almost certainly not be successful. In research labs, stakeholders include lab technicians and scientists as well as line-of-business managers and analysts in the finance department.
The CoE should review and approve proof of concept (PoC) projects, develop KPIs, evaluate the success of PoCs and monitor ongoing Iot deployments.
As an Iot project rolls out, it is crucial that the CoE focus on business value and not on the technology. An organization needs to be ready to abandon any Iot project at key stages if KPIs cannot be developed or met, the PoC fails or business targets are missed.
Internet of Things platform
A collection of Iot enabled devices can quickly turn into a collection of data silos, which in turn are a collection of data challenges. For your Iot deployments to be successful, the data needs to be aggregated and searchable. To maximize value, data should flow into a centralized things platform, where they are aggregated, correlated and analyzed with scientific study or operations data.
While the Iot industry is still in its infancy, some laboratory informatics vendors have begun to integrate Iot data into their study management platforms. As you plan your Iot future, we recommend looking carefully into where your data are streaming and how that data will be integrated into your informatics ecosystem.
Monitoring ioT data streams
Iot data streams reflect the current state of the physical world and thus have real-time value. Anomaly detection algorithms can monitor IoT data streams and send alerts when significant changes or off-nominal trends are detected.
Some examples where anomaly detection algorithms can prove useful in animal facilities include:
* Trends: Suppose the humidity of an animal cage is gradually increasing? This might portend a failure in the air handling system or be caused by a leaky water bottle.
* Dynamic range: Suppose you are monitoring noise levels in an animal cage? If noise levels suddenly oscillate beyond normal range, it might indicate that animals are fighting.
* Spike detection: A sudden spike of light in a cage might indicate that the light control in an animal room is not functioning as expected.
IoT is at a level of technical maturity where it is poised to have transformative impacts on animal care and research. Technology adopters should establish an IoT CoE and develop business cases before investing in large scale projects. In order to maximize science and business value from IoT, data need to be delivered to a "things" platform, where they can be correlated and analyzed in context with science and operations data.
By Chuck Donelly and Julie Morrison, RockStep Solutions, Inc.
|Printer friendly Cite/link Email Feedback|
|Author:||Donelly, Chuck; Morrison, Julie|
|Date:||Apr 1, 2019|
|Previous Article:||Mutant Gene Controls How Stem Cells Become Skin Cells.|
|Next Article:||Marine Genomics Aids Human Disease Research: The Gloucester Marine Genomics Institute is dedicated to decoding the mysteries of sea life DNA.|