Date posted: January 9, 2019
Tom Stein
There is no argument new smart technologies are coming in swine production says Tom Stein, senior strategic adviser for Maximus Systems. Here are nine examples already here or on the way.
Real-time activity and behavior monitoring to detect problems in pigs or poultry. Changes in pig eating and drinking patterns can be used to detect sickness and provide an early warning system. Compared with trained observers, the data could detect sick pigs at least 24 hours before a human walking through the barn would be able to.
Real-time behavior monitoring to detect estrus in sows. PigWatch (Ro-main; Quebec, Canada) technology uses infrared sensors mounted above a gestation stall to monitor and track sow activity and behavior. Using a machine learning algorithm, it determines not only if and when a sow is in estrus but also the optimal time for insemination.
Real-time audio signal processing to detect respiratory problems in pigs. The Pig Cough Monitor (Soundtalks NV, Belgium) uses a microphone hanging over pens of finishing pigs and a machine learning algorithm to identify and make sense of pig coughing events. It is the most robustly and formally studied audio processing sensor so far. European farm research trials over the past 10 years demonstrated the tool's potential to deliver early warnings up to two weeks earlier than a situation where producers or veterinarians rely on their own routine observations without the monitor. Automated cough detection and monitoring can detect different causes of respiratory problems as well as technical problems with ventilation.
Real-time audio signal processing to prevent crushing deaths in neonatal piglets. SwineTech (Cedar Rapids, Iowa) is a company aiming to prevent piglet deaths due to crushing by sows within the first four days after farrowing. The ECHO hardware device listens for piglet sounds and squeals. When it identifies the unique sound of a piglet being crushed, it sends a vibration to a wearable patch attached to the sow. The vibration causes the sow to immediately stand up which prevents a baby pig from being crushed. Several large pork producers in the Midwest are doing field trials and early results show a statistically significant reduction in crushing deaths (70 percent lower than controls) and an increase in Day 4 livability by 0.5 to 1.0 pigs per litter. Researchers at Kansas State showed that the ECHO system was more effective than a conventional human hand-slap at causing sows to stand up while causing a sow no more stress than the conventional method.
Voice-enabled data entry and workflow. AgVoice (TekWear, LLC; Atlanta, Georgia) is a hands-free data entry and workflow management service that helps agriculture professionals capture insights while on the move. Users include agronomists, plant scientists, veterinarians and machinery repair specialists, among others. The hands-free, voice-interactive technology has the potential to fundamentally shift the way food and agriculture professionals capture usable insights and data.
Real-time image processing to weigh pigs. Researchers at the French Pig & Pork Institute in France have developed a 3D imaging prototype for weighing individual pigs in partnership with Advansee, a company specialized in image analysis and embedded electronics For weight estimates, the objective is for a margin of error between 2 percent and 4 percent. The EyeGrow system (Fancom BV, the Netherlands) has recently been introduced. It uses a 3D infrared camera and machine learning to weigh pens of finishing pigs. It produces average daily weight gain for the pen along with variation in weights measured by standard deviation. They estimate the cost at $1,400 USD per camera with one camera needed for each pen of 20 finishing pigs.
Tracking movement of on-farm workers using Bluetooth to improve biosecurity. A new system has been tested using small devices, called beacons, which are worn by farm staff. A beacon is a small Bluetooth low energy radio transmitter and can be used as an indoor positioning system for farm staff inside the barn. Three types of staff movements are collected: (1) Correct. From higher health barns to lower (i.e., from gestation to lactation barn); (2) Wrong. Low risk. From lower health barns to higher (i.e., from lactation to gestation); (3) Wrong. High risk. From lower health barns to higher (i.e., from finisher barns to gilts).
Geofence-based feed traceability using real-time GPS location data. FeedTrackur (DP Techlink, LLC; Des Moines, Iowa) is a hardware-software solution for automated and real-time traceability of animal feed. A cloud-based database contains continuously updated feed orders, load-out information, driver information, feed truck information, and real-time GPS data which enables location tracking. Cloud-based mapping software contains GPS-coordinates for all physical locations including sites, barns, and feed bins. Each is demarcated with geofences based on specific GPS coordinates. Feed truck drivers interact with a tablet computer that sits inside the cab. Whenever a feed truck crosses a geofence, logic is triggered as to whether it is the correct location for the feed delivery and, if it is, the delivery location (the feed bin itself) is GPS-stamped for traceability. If it's not the correct location, the software alerts the driver and prevents mistakes.
Smart digital monitoring of medical treatments using RFID, Bluetooth, and Wi-Fi. Automed (Canberra, Australia) and V-ETIC (Henke-Sass Wolf GmbH; Tuttlingen, Germany) have each recently introduced Wi-Fi and Bluetooth-enabled smart dosing devices that provide automatic medication delivery, data recording, and weight-based and fixed-amount dosing. Both systems record treatments, manage medication inventory, and integrate with scales, RIFD readers and with livestock management software. Both have apps that allow users to control, configure and manage the dosing devices. Because they track medication inventory and real-time usage, they enable compliance and traceability between drug inventory and use in animals.