A Real-Time Automated System for Monitoring Individual Feed Intake and Body Weight of Group-Housed Young Chickens
, Zhixiong Zeng 1,2,*, Enli Lv 1,3, Xinyuan He 1, Boyang Zeng 1, Fan Wu 1, Jiaming Guo 1 and Ziwei Li 4
Round 1
Reviewer 1 Report
Why and how 30 grams is selected as the typical feed weight in the feeding tray? How this study helps in identifying, the optimal feed rate(per day) for chicken? How is overfeeding avoided using the automated system? How to interpret the figure 12? Is this data obtained from one chicken ? or all chicken ? What is the reason behind the abnormal data? How many chickens fall under this outlier data and what would be the next steps? Is there any feeding pattern identified between the healthy chicks and unhealthy chicks? How the automatic machine will know when to stop feeding the chicks during the day? The Sentence formations can be drastically improved. How this work can be included in the real world and what would be the challenges?
Author Response
Point 1: Why and how 30 grams is selected as the typical feed weight in the feeding tray?
Response 1:
We mainly obtained through the preliminary raising chickens' experience that the feed intake of yellow broilers weighing 50-1500g is less than 30g each time; We also refer to the literature “Relationships between feeding behaviors and performance traits in slow-growing yellow broilers”. In this literature, the feed intake of yellow broilers with a body weight > 2200g is less than 50g each time. Because the yellow broilers in this paper are younger and their body weight is less than half of the 2200g mentioned in the literature, so 30g is used as the feed refill target for the tray is reasonable.
Point 2: How this study helps in identifying, the optimal feed rate (per day) for chicken?
Response 2:
The optimal feed rate ( per day ) for chicken is determined by the number of birds fed by each system. The total feeding time of the system is fixed (from 5:00-19:30) per day, so if we make sure that every bird has the opportunity to eat, the daily feeding time of each chicken can be roughly obtained from the total time divided by the number of chickens; Our system feeds chickens ad libitum that if there are no chickens on the weight scale, any chicken can enter and eat, and only the bird leave the weight scale, other chickens can enter, and the direction of leaving is different from the entering, which makes the chicken that just left far away from the entrance of the passage than others, that ensures the chicken currently leaving at least have to wait until the next time to enter again. And those who have not eaten will have a stronger desire than those who have eaten to enter the passage, so the possibility is greater than for those who have already eaten. Generally, the probability of feeding each chicken is equal; We also found that the chicken that eats the most every day in the flock grows the fastest, but the FCR (feed/meat) is not the best, so we can choose the daily feeding time of the chicken with the best FCR in the flock used as a reference to determine the number of chickens fed by each system, to obtain the optimal daily feeding rate.
Point 3: How is overfeeding avoided using the automated system?
Response 3:
Maintain the number of chickens fed by each system to a reasonable number, and ensure that each chicken in the flock has an equal opportunity to eat in sequence, so as to limit the total feeding time of each chicken per day, which is the method of avoided overfeeding by the automatic system. To ensure that every chicken in the flock has an equal opportunity to eat, our system is to refill the tray only when the weight scale is idle, and the single passage enables separated the leaving chicken from the weight scale to prevent returning, and the chicken just leaving feed zone are farther from the entrance than those who waiting outside, and has a low appetite to eat which make chickens eating in sequence. Chickens will leave the scale due to an empty tray, water temptation, other chickens' calls, or driving away when they eating. In this paper, we mainly through the ways that the chicken has eaten up or need to drink water to leave, and we have obtained some achievement on driving away chickens, which will be applied to this system to drive away the reluctant chickens after feeding.
Point 4: How to interpret the figure 12? Is this data obtained from one chicken? or all chicken? Response 4:
Figure 12 shows the body weight of a single chicken within 40 days. This data is retrieved from the database which is permanently stored in the HMI, and it includes all data of group-housed chickens during the experiment. For each feeding, the body weight, the number of RFID tags, feed amount, drop amount, feeding duration, and the end moment will be recorded in the database, so the body weight can be retrieved through the numerical of the RFID tag from the database for any chicken within the feeding period.
Point 5: （figure 12）What is the reason behind the abnormal data? How many chickens fall under this outlier data and what would be the next steps?
Response 5:
The abnormal data (Figure 12) refers to the data that cannot represent the real body weight. For example, the weight record is much smaller than the average weight of the day, and the abnormal data may be due to part of the chicken's body being supported by the scale, it could deem incomplete-weighing, for instance, one foot of the chicken is on the rack and the other is on the scale, so all chickens may have this kind of data anomaly. In this paper, the median mean filter is used to calculate the real weight of the chicken, the weight sequence is sorted from small to large and the average value of the middle part is taken as the real weight. The incomplete-weighing sequence will generally be shaved after data processing, so the limited-time incomplete-weighing has no effect, when the abnormality occurs the whole time, but the position of the chicken cannot eat the feed, it has no effect on the feed intake, and it is easy to distinguish as the abnormal data are much smaller than the average body weight of the day, so it can be ignored too, and the fundamental solution is to develop a driving-away mechanism to directly drive the chickens away while they have not eaten for a long time.
Point 6: Is there any feeding pattern identified between the healthy chicks and unhealthy chicks?
Response 6:
There is no model to distinguish between healthy and unhealthy chickens currently. So we will further study the differences in the feeding behavior of chickens in different conditions to establish a feeding model.
Point 7: How the automatic machine will know when to stop feeding the chicks during the day?
Response 7:
The feeding time is determined by the single passage in this paper. We set the timing in the controller according to the activity of the chicken. After 19:30, the door of the single passage can only be opened from the inside of the feeding zone, so when the last chicken comes out, the day's feeding is over, and at 5:00 in the morning, the restriction is lifted, and the chickens can go to the feeding zone to eat.
Point 8: The Sentence formations can be drastically improved.
Response 8:
Thank you for your suggestions. Our paper will be further grammatically revised through the paid revision service provided by MDPI. If you feel that there are still sentence formation shortcomings that need improvement, please don't hesitate to contact us. (https://www.mdpi.com/authors/english )
Point 9: How this work can be included in the real world and what would be the challenges?
Response 9:
Our research on the system of auto-measure feeding behavior is always ongoing. The biggest challenges of the system in previous research are the three aspects mentioned in the paper. First, it was the low accuracy of dynamic body weight weighing; Second, the chicken's competition for food did not guarantee that the measurement is only one chicken; Third, the amount of feed reduction in the tray is not exactly related to the actual feed intake. In the paper, we have solved the problems, but according to the feedback, the feed drop weighting scale will bring extra work to the worker when manual cleaning is required after the drop amount has accumulated to a big amount. To solve this we also have developed an innovative way to measure the actual feed intake. And another challenge is the way of driving away chickens, especially young ones are more likely to stay in the feeding zone after eating. Now, we have also made some breakthroughs on driving away chickens, we will show it in soon.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The authors provided an area of timely and important scholarship given the necessity to improve food security around the world.
The manuscript does an adequate job of communicating the findings to current and future researchers.
The gap that needs filling is the role this scholarship informs pracititioners or informs change agents (Extension, etc.) in disseminating information to stakeholders for, hopefully, their adoption and resulting impacts in their poultry production endeavors. To fill this gap, it is recommended to expand the literature review and discussion sections, with these citations to improve how this scholarship informs practice and therefore, increase the adoption of a real-time automated system:
Agricultural 4.0 and Industry 4.0 technologies -
https://doi.org/10.3390/s22186833
Food security and agricultural challenges -
https://doi.org/10.1080/10942912.2022.2066124
All of these two would strengthen the introduction respective to agricultural innovations impact and advancement of agricultural stakeholders and agricultural organizations/systems but more importantly help connect your findings to the potential impacts of real-time automated systems' adoption in the improvement of global food security challenges. This is how you can demonstrated your research (you) can improve practice (farmers).
The above provided citations not only add to your introduction but make a stronger connection and case in the discussion section of impact for who, how, and why it matters. These additions and connections also improve the long-term sustainability (increased readership and potential resulting increase citations) of your scholarship.
Author Response
Point 1: The gap that needs filling is the role this scholarship informs practitioners or informs change agents (Extension, etc.) in disseminating information to stakeholders for, hopefully, their adoption and resulting impacts in their poultry production endeavors. To fill this gap, it is recommended to expand the literature review and discussion sections, with these citations to improve how this scholarship informs practice and therefore, increase the adoption of a real-time automated system:
Agricultural 4.0 and Industry 4.0 technologies -https://doi.org/10.3390/s22186833
Food security and agricultural challenges- https://doi.org/10.1080/10942912.2022.2066124
All of these two would strengthen the introduction respective to agricultural innovation's impact and advancement of agricultural stakeholders and agricultural organizations/systems but more importantly help connect your findings to the potential impacts of real-time automated systems' adoption in the improvement of global food security challenges. This is how you can demonstrate your research (you) can improve practice (farmers).
The above provided citations not only add to your introduction but make a stronger connection and case in the discussion section of impact for who, how, and why it matters. These additions and connections also improve the long-term sustainability (increased readership and potential resulting increase citations) of your scholarship.
Response 1:
We added the 2 references you recommended in line 43 of the paper, as you said, it makes the paper a stronger connection and case in the discussion section of impact for who, how, and why it matters.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
The manuscript is much improved.
Opportunities to convey the research findings to practitioners (poultry farmers and leaders within global poultry industries) is still not adequate in the Conclusions - in my opinion.
The authors stated on page 12 (lines 396-398) "The results suggest that we have improved the accuracy of live poultry weighing the real-time automated monitoring system developed in this study, paired with a single passage, can accurately monitor the body weight changes and feeding behavior of young chickens". This is a critical positive implication of the research. The connection, importance, and the extent the research informs farmers and the industry mentioned above is missing in the scholarship.
Therefore, I believe the Conclusions needs expansion delineating the imperative nature of the findings to the end users (farmers, industry, future researchers, extension or change agents working in communities, etc.).
What good is our research if we cannot communicate it to stakeholders that may, hopefully, impact their farming operation, community, and society in general?
Again, the manuscript is much improved. The importance for and implications concerning poultry farmers and the poultry industry is missing and therefore, hinders the publication of this scholarship in this reviewer's opinion.
Author Response
Point 1:
Opportunities to convey the research findings to practitioners (poultry farmers and leaders within global poultry industries) is still not adequate in the Conclusions - in my opinion.
The authors stated on page 12 (lines 396-398) "The results suggest that we have improved the accuracy of live poultry weighing the real-time automated monitoring system developed in this study, paired with a single passage, can accurately monitor the body weight changes and feeding behavior of young chickens". This is a critical positive implication of the research. The connection, importance, and the extent the research informs farmers and the industry mentioned above is missing in the scholarship.
Therefore, I believe the Conclusions needs expansion delineating the imperative nature of the findings to the end users (farmers, industry, future researchers, extension or change agents working in communities, etc.).
What good is our research if we cannot communicate it to stakeholders that may, hopefully, impact their farming operation, community, and society in general?
Again, the manuscript is much improved. The importance for and implications concerning poultry farmers and the poultry industry is missing and therefore, hinders the publication of this scholarship in this reviewer's opinion.
Response 1:
As you said, we have connected poultry producers with our study in the conclusions, greatly appreciated your suggestions, and do not hesitate to contact us if there are still gaps. The revised conclusions are as follows:
The results suggest that we have improved the accuracy of live poultry weighing, and the real-time automated monitoring system developed in this study, paired with a single passage, can accurately monitor the body weight changes and feeding behavior of young chickens. Our research may be useful in poultry breeding or used as a verification device for other breeding methods, it could improve the growth rate of poultry and make the FCR better, which would reduce the time of raising poultry and improve the feed utilization rate, shorten the production cycle and reduce the feeding cost. It is beneficial to alleviate the current food crisis facing the world. Our research is also of great significance for poultry weight tracking and feeding behaviors monitoring, which would improve the automation level of poultry production, and timely detection of abnormal behavior in poultry to prevent major economic losses before they occur, bringing economic benefits for poultry farmers.
