Count unique values with criteria and blanks and an average of maximum consecutive values

I notice that times and serials are not contiguous. Are you tracking more than one day in this data? If so, then you also need a date. If not, then the data needs to be sorted by serial and time.

Here is my shot at it assuming that it is all in one day.

First I sorted by Serial and date so we can get the most consecutive values lined up.

Then I introduced a bunch of helper columns:
- shift =IF([@Time]>=0.5,"PM","AM") - 0.5 days is one half day or 12 hours so 0.5 is a shorthand way of telling Excel, Noon.
- composite =[@Serial]&":"&[@shift] - the combination of serial and shift make a record unqiue
- Unique =MATCH([@Composite],[Composite],0)=ROW()-1 - this formula is true for the first occurrence of a composite.
- Run =IF([@Unique]=TRUE,1,G1+1) - this is a running total of the number of records in the "run" of Serial + Time
- Max =MAXIFS([Run],[Composite],[@Composite]) - gets the maximum number in a run.

Then there are the results
Average Change (Total) =AVERAGE(Table1[MAX])
Average Change (PM) =AVERAGEIFS(Table1[MAX],Table1[shift],"PM")
Average Change(AM) =AVERAGEIFS(Table1[MAX],Table1[shift],"AM")

Unique Values (PM) =COUNTIFS(Table1[Unique],TRUE,Table1[shift],"PM")
Unique Values (AM) =COUNTIFS(Table1[Unique],TRUE,Table1[shift],"AM")